View Article PDF

Infective endocarditis (IE) is a complex condition to diagnose and manage that continues to cause significant morbidity and mortality in New Zealand and internationally.[[1–5]] There is a trend towards an increasing proportion of IE cases caused by Staphylococcus aureus in high-income countries, while streptococcal infections continue to predominate in lower-income settings.[[2,6,7]] The International Collaboration on Endocarditis-Prospective Cohort Study (ICE-PCS) is the largest prospective cohort study on IE to date, collecting data on 2,781 patients from 58 hospitals across 25 countries (including New Zealand) between 2000 and 2005.[[2]] The ratio of streptococcal to staphylococcal species causing IE was 0.7. Other studies have demonstrated ratios of 2.5 in Pakistan and South Africa, 1.6 in South America, 1.2 in India, 0.8 in Europe and 0.3 in North America.[[5]] The higher rate of streptococcal IE in lower-income countries has been attributed to poorer dental health, higher prevalence of predisposing rheumatic heart disease (RHD), and that intravenous drug use (IVDU) and healthcare-associated risk factors for staphylococcal IE are less common.[[5,8]]

Several previous studies have explored the demographics, clinical features and management of IE in New Zealand.[[1,3,4,9]] The largest was the New Zealand component of ICE-PCS, which included 337 cases of IE.[[3]] There is a paucity of IE incidence data internationally and in New Zealand.[[5]] A systematic review of population-based studies from 1969 to 2000 across seven high-income countries reported an incidence of 1.4 to 6.2 cases per 100,000 persons.[[10]] Other studies in Hong Kong, Italy and Australia found an incidence of 2.8, 4.4 and 4.7 per 100,000 person-years respectively.[[11–13]] No previous New Zealand studies have provided population-based IE incidence data or estimated the healthcare cost of IE.

Dental disease can predispose to IE through bacteraemia from oral flora.[[14,15]] There are inequities in dental health in New Zealand, with higher rates of dental disease observed in the Māori population and among families of lower socio-economic status.[[16–19]] The Northland Region has some of the highest rates of social deprivation and dental disease in the country.[[18–20]] Dental care for adults in Northland is predominantly through private practice, other than limited emergency dentistry services for medically compromised and low-income adults in hospitals and community clinics.[[21]] There is no fluoridation of the water supply in Northland.[[21]] We hypothesise that the high burden of dental disease in Northland may be contributing to a high incidence of IE from dental organisms such as oral streptococci and enterococci.

RHD, a complication of acute rheumatic fever (ARF), is a condition associated with poverty and overcrowded living conditions that is now rarely encountered in most high-income countries. However, ARF and RHD present an ongoing challenge in Northland. A review of ARF in Northland found an incidence of 7.0 per 100,000-person-years, with 93% of cases occurring in Māori and 87% of patients living in deprived areas.[[22]] No previous studies have explored the association between RHD and IE in Northland.

This study aimed to define the epidemiology, risk factors, microbiology, presentation, management and healthcare impact of IE in Northland, to guide strategies for prevention and quality improvement.

Method

Setting

The Northland Region is defined in this study based on the 2018 census map boundaries.[[23]]Northland District Health Board (NDHB) includes five public hospitals, located in Whangārei, Bay of Islands, Kaitaia, Hokianga and Dargaville. All patients fulfilling the Modified Duke Criteria for IE were eligible for inclusion if they received treatment at any of these sites between 1 January 2010 and 31 December 2019. There is no inpatient cardiothoracic surgical service within NDHB. Northland patients requiring cardiothoracic surgical input are routinely transferred to Auckland District Health Board (ADHB).

Study design and data collection

Cases of IE were identified retrospectively using discharge coding data from NDHB hospitals (International Classification of Diseases-10 codes 133, 138, 139). Cases of all ages were included in the study if they met Modified Duke Criteria for ‘definite’ or ‘possible’ IE.[[24]] Demographic and clinical data were extracted from electronic medical records and entered on an audit tool (Microsoft® Excel (version 16.32)). New Zealand Index of Deprivation 2018 (NZDep2018) deciles were assigned according to each patient’s address at diagnosis. Territorial authority boundaries from Stats NZ were used to define the Far North, Whangārei and Kaipara districts. Nosocomial IE was defined as IE developing in a patient hospitalised for more than 48 hours prior to onset of signs or symptoms of IE.[[2]] Dental information was extracted from patients’ medical notes and the Titanium® dataset, an information management system that records clinical data from free dental examinations conducted by dental therapists and dentists.[[19]] Costing data were obtained from the NDHB webPAS® portal and included all admission costs at NDHB and ADHB for management of IE and its complications, inter-hospital transfers, cardiology clinics within 12 months after discharge and community nursing visits within 31 days of discharge.

Data analysis

As population growth occurred in Northland during the study period, population data from the approximate midpoint of the study (the mean of 2013 and 2018 census data) were used for incidence calculations. The incidence of IE was determined by dividing the number of incident cases by the total number of person-years accumulated in the study population. Incidence values and 95% confidence intervals were calculated using Microsoft Excel. Other statistical analyses were performed using IBM® SPSS Statistics (version 25.0.0.0). Continuous variables were reported as medians and interquartile ranges. Categorical variables were expressed as frequencies and percentages of the specified group. Categorical data were compared using the Chi-squared test or Fisher’s exact test. Parametric data were compared using the t-test or ANOVA test and non-parametric data using the Mann–Whitney or Kruskal–Wallis test. The Pearson correlation coefficient was used to calculate the correlation between two continuous variables.

Ethics approval

Ethics approval was sought from the New Zealand Health and Disability Ethics Committees. The study was deemed out of scope and not requiring ethics review.

Results

Incidence and demographics

The Northland population estimate used for incidence calculations was 165,384 and the Māori population was 56,807 (34%). Census data indicate 18% population growth in Northland from 2013 to 2018.[[23]] Between January 2010 and December 2019, there were 140 episodes of IE identified in 134 patients. Ninety-seven cases (69%) were classified as definite and 43 (31%) as possible IE. Table 1 describes their demographic characteristics.

The overall incidence of IE in Northland was 8.47 (95% CI 7.12–9.99)per 100,000 person-years. The incidence in males was 12.82 per 100,000 person-years,compared to 4.27 in females (p=0.04). The incidence in the New Zealand European(NZE) population was 8.30 and in Māori was 6.51 (p=0.64). Figure 1 illustrates the incidence stratified by age and ethnicity. Thepopulation group at highest risk of IE were Māori aged 80–84 years,with an incidence of 72.90 cases per 100,000 person-years compared to 18.77 amongnon-Māori in the same age group (p <0.001).Rates of IE were also significantly higher among Māori aged 25–29, 30–34,45–49 and 75–79 years.

View Table 1 & Figure 1.

One hundred and two patients (73%) lived in areas with NZDep2018 deciles between 7 and 10 (Table 1). A higher proportion of Māori than non-Māori patients lived in decile 9 and 10 areas (58% versus 39%). Figure 1 illustrates each patient’s domicile at the time of diagnosis.

Risk factors

There were 62 cases of prosthetic valve endocarditis (PVE) (44%) and 78 cases of native-valve endocarditis (56%). In 17 of the PVE cases (27%), the valve had been replaced due to RHD. Twenty-five patients with PVE (40%) had mechanical valves and 37 (60%) had bioprosthetic valves. A history of congenital valve disease was documented in 25 patients (18%) and RHD in 19 patients (14%). There were 13 cases of recurrent IE, of which six were managed for IE twice at NDHB during the study period and seven had a history of IE prior to the study period. Five cases (3%) were classified as nosocomial. Three patients (2%) had a history of injecting drug use. Further patient co-morbidity data can be found in Table 1.

Microbiology

Table 2 outlines the microbiology of IE cases according to valve type. Viridans streptococci were the most common causative organism (n=42, 30%), followed by S. aureus (n=32, 23%) and E. faecalis (n=22, 16%). The ratio of streptococcal to staphylococcal species causing IE was 1.7.

The five cases of nosocomial IE were caused by methicillin-sensitive S. aureus (MSSA) (n=1), viridans streptococci (n=1), coagulase negative staphylococci (n=1), non-HACEK group gram-negative bacilli (n=1) and polymicrobial IE (n=1).

View Table 2.

Clinical presentation

Table 3 describes clinical and investigation findings. The most frequent clinical signs were fever (n=116, 83%), heart murmur (n=92, 66%) and splinter haemorrhages (n=18, 13%). Other classical signs of IE were rare, with few patients exhibiting Janeway lesions (n=2, 1%), Roth spots (n=2, 1%) or Osler’s nodes (n=1, 1%).

In 135 cases (96%) the C-reactive protein (CRP) level was 10mg/L or higher on admission. The median CRP level on admission was 89 mg/L (IQR 52–155).

Blood cultures were positive in 129 cases (92%). Of the 11 cases with no positive blood cultures, eight had received antibiotic therapy prior to cultures being taken. The median number of blood cultures taken per patient was five (IQR 4–8), and the median number of positive blood cultures was two (IQR 1–3).

Transthoracic echocardiography (TTE) was performed in 131 patients (94%) and vegetations were identified in 48 of these (37%). Transoesophageal echocardiography (TOE) was performed in 107 patients (76%) and vegetations were identified in 77 (72%). One hundred patients (71%) underwent both TTE and TOE. Vegetations were identified on both TTE and TOE in 29 of these patients (29%) and vegetations were seen on TOE but not TTE in 48 (48%). The sensitivity of TTE compared to the gold standard of TOE in detecting valvular vegetations was 38% and the specificity was 87%. In cases of pure aortic valve IE the sensitivity of TTE compared to TOE was 45%. In cases of pure mitral valve IE the sensitivity of TTE compared to TOE was 30%. There was no significant difference in rates of TOE by ethnicity (n=25, 72% in Māori; n=81, 78% in non-Māori; p=0.490), or territorial authority (n=33, 87% in Whangārei; n=26, 76% in Far North; n=10, 91% in Kaipara District; p=0.382).

The aortic valve was involved in 79 cases (56%), mitral valve in 49 (35%), tricuspid valve in eight (6%) and pulmonary valve in four (3%). In 12 cases (9%) there were multiple valves affected. In 14 cases (10%) the affected valve(s) were unknown.

View Table 3.

Dental history

Ten cases (7%) had a recorded public dental service review in the 12 months preceding diagnosis of IE, of whom six had pre-existing prosthetic valves. Only one of the 19 cases with a history of RHD had a recorded dental review in the 12 months prior to IE diagnosis (5%). Of the 42 cases of viridans streptococcal IE, five had a recorded dental review in the prior 12 months (12%). No patients with IE caused by other streptococci or E. faecalis had a recorded dental review in the prior 12 months.

Nineteen cases (14%) underwent dental review during their hospital admission, including 13 of those with viridans streptococcal IE (31%), zero with other streptococcal IE, zero with E. faecalis IE and three with S. aureus IE. Twenty-one cases (15%) had a dental x-ray during their admission.

Complications

Complications of IE included systemic embolism (n=53, 38%), congestive heart failure (n=42, 30%), stroke (n=33, 24%), paravalvular abscess (n=19, 14%), valvular perforation (n=16, 11%), chordal vegetation or rupture (n=15, 11%), persistent bacteraemia for more than two weeks (n=10, 7%) and death within six weeks of diagnosis (n=16, 21%). There were 14 cases of prosthetic valve regurgitation (23% of all PVE cases) and 11 cases of prosthetic valve dehiscence (18% of PVE cases). An overview of complications stratified by causative organism is provided in the Table 2.

Management and outcomes

Valve surgery was performed in 45 cases (32%). Aortic valve surgery was performed in 23 patients (16%), mitral valve surgery in 26 (19%), tricuspid valve surgery in four (3%) and pulmonary valve surgery in two (1%). In 10 cases (7%), multiple valves required intervention. Further detail on types of valve surgery is provided in the Table 3. The median time from initial contact with healthcare services to valve surgery was 24 days (IQR 10–58). Valve surgery was required in 41% of S. aureus cases (n=13), 26% for viridans streptococci (n=11), 18% for E. faecalis (n=4) and 66% for coagulase negative staphylococci (n=2). Nine patients (6%) required non-cardiac surgery. Peripherally inserted central catheters were placed for 115 patients (82%) and 64 patients (46%) received outpatient intravenous antibiotic therapy.

The all-cause mortality rate at six weeks after diagnosis of IE was 18% (n=25), at six months was 22% (n=31), 12 months was 25% (n=35) and five years (or duration of follow up if less than five years) was 36% (n=51). Among patients who underwent valve surgery the six-week mortality rate was 18% (n=9), which was equivalent to the non-operative group (18%, n=16, p=0.974). In those who received ICU-level care the mortality rate was 25% (n=14), compared to 13% (n=11) in those not admitted to ICU (p=0.07). There was no significant difference in rates of ICU admission, valve surgery or death according to patients’ home territorial authority (p=0.499, 0.353 and 0.110 respectively).

Healthcare impact

Table 4 outlines the healthcare impact of IE and sites of inpatient treatment. An estimated NZ$6,560,470 was spent on direct patient care for IE during the study period. The median healthcare cost per case of IE was $34,053 (IQR $18,212–69,994).

For streptococcal IE the median cost per case was $29,362, for S. aureus was $40,420 and for E. faecalis was $30,265 (p=0.278). Length of hospital stay was directly associated with healthcare cost (correlation coefficient 0.584, p<0.000). At least one inter-hospital transfer was required for 96 of the cases (69%; range 0–6 transfers).

View Table 4.

Discussion

This study describes the burden of IE in Northland, identifying the elderly Māori population as being at greatest risk. The overall IE incidence of 8.47 per 100,000 person-years in Northland is higher than has been described in other high-income countries (reported incidence of 1.4 to 6.2 per 100,000 person-years).[[10–13]] No previous New Zealand incidence data are available for comparison.

The severity of IE cases in Northland is notable, with higher rates of systemic embolism (44% versus 17%), stroke (27% versus 14%) and death (21% versus 6%) than were observed in the New Zealand ICE-PCS cohort.[[3]] This is despite ICE-PCS only enrolling patients from large hospital centres with a cardiothoracic surgical service, which was expected to select out cases that are more complex and requiring surgical intervention.[[3]] However, the rate of valve surgery in Northland cases was similar to in the ICE-PCS cohort (34% and 33% respectively). There was a markedly longer duration from initial healthcare contact to surgery in Northland cases than in ICE-PCS (median of 24 days versus 4 days). This may be due to delays in diagnosis and inter-hospital transfer from NDHB to a cardiothoracic surgical centre, or differences in the pathology and acuity of presentation. Ethnicity and deprivation were not described for the ICE-PCS cohort. However, the rate of RHD, an indicator of underlying poverty and overcrowding, was much higher in our study (14% versus 4%) and likely reflects inequities in healthcare access and social determinants of health in Northland.

The microbiology of IE in Northland is more consistent with that seen in lower-income countries, with a high proportion of cases caused by streptococci rather than staphylococci (ratio 1.7 in Northland, versus 0.7 in the international ICE-PCS study). This pattern supports our hypothesis of dental disease in Northland contributing to the high burden of IE.

Factors contributing to the higher incidence of IE among those aged over 70 years, particularly Māori, may include a greater prevalence of cardiac valvular disease, dental decay, diabetes and immunocompromising medical conditions within this population. Changes in dental profiles with increasing retention of natural teeth may also predispose to higher rates of dental infection in later life.[[24]] Elderly people may also have difficulty accessing dental care (both for routine prevention and acute dental infections), due to disability or financial constraints. These inequities may be more pronounced among elderly Māori than non-Māori due to cultural and language barriers, remoteness and higher rates of social deprivation.[[20]]

This study also highlights the poor sensitivity of TTE (38%) compared to the gold standard of TOE in the diagnosis of IE. A systematic review and meta-analysis published in 2017 found similar results, with fundamental TTE detecting vegetations with a sensitivity of 41% (95% CI 29–55%) compared to the reference standard of TOE.[[25]] We hypothesise that reduced TOE availability in Northland may have contributed to delayed diagnosis and valve intervention. Within our cohort there was no significant difference in rates of TOE by patient ethnicity or region. However, it would be useful to compare the proportion of patients undergoing TOE and time from admission to TOE for evaluation of endocarditis between district health boards, to assess equity of access at a regional level.

It is possible that cases of IE were missed in this study if no discharge summary was completed (required for hospital coding) or the diagnosis of IE was not included during coding. Therefore, our incidence data may underestimate the true burden of IE. Another limitation of incidence calculations was the use of midpoint population estimates, as significant population growth occurred over the study period. Co-morbidity results and clinical features may be underestimated if medical history or examination was incompletely documented or not included in patients’ electronic admission or discharge records.

Only public dental data from free dental examinations were accessed. As dental care for adults in Northland is predominantly through private practice, most dental encounters are likely to have been missed (unless documented in patient notes). However, data on inpatient dental reviews at Whangārei Hospital are expected to be complete as these are routinely performed via the on-site public dental service and recorded on the Titanium database.

If patients were transferred to or from a district health board other than NDHB or ADHB (or overseas), then data from these other sites were not accessible. Assessment of longer-term outcomes such as delayed surgery, recurrence and one- and five-year mortality was not available for all patients as the time had not yet elapsed at the time of data collection.

Cases were included in the cohort if they met Modified Duke Criteria for ‘possible IE’. Some of these cases may have been incorrectly diagnosed as IE (such as episodes of bacteraemia without endocarditis). However, cases were excluded if they met Modified Duke Criteria for ‘rejected IE’ or if IE was not documented as a discharge diagnosis by the treating team.[[26]]

The calculated total healthcare cost of IE during the study period is likely a significant underestimate as it does not include non-cardiology outpatient clinic follow-up, private healthcare encounters, or the cost of long-term patient disability, interruption to employment and years of life lost. Further studies measuring the burden of disease from IE, including years of life lost and years lived with disability, would be valuable in quantifying the broader societal cost of the disease.

As a high proportion of IE cases were caused by odontogenic organisms, a review of factors contributing to dental disease and further investment in oral health promotion in Northland is recommended. The New Zealand Promoting Oral Health guideline provides a comprehensive framework for based on Ottawa Charter principles.[[27]] Community fluoridation of reticulated water supplies supplying populations over 500 has been assessed as a highly cost-effective strategy for prevention of dental caries in New Zealand.[[28]] Research indicates that populations living in deprived areas may show the greatest reduction in dental ambulatory sensitive hospitalisations as a result of community water fluoridation.[[29]] At present, no reticulated water supplies in Northland are fluoridated.[[30]] Consideration should be given to water fluoridation for dental protection in Northland, a position which is supported by NDHB.[[28,30]]

There is limited evidence that oral health education alone can improve dental hygiene.[[31]] However, a multi-component approach including oral health promotion in schools, provision of fluoride-containing toothpaste, oral health training of non-dental professionals and limitations on marketing and sale of high-sugar food and beverages may be beneficial.[[31]] Further study is currently underway to develop and evaluate interventions to reduce oral health disparities for Māori.[[32]]

Previous research suggests that improvement in oral hygiene by regular dental scaling is associated with a significant decrease in risk of IE.[[33]] Access to affordable dental care is particularly important for high-risk groups such as elderly Māori and patients with prosthetic valves, RHD, congenital valve disease or previous IE. International guidelines recommend that high-risk patients undergo dental follow-up twice a year.[[34,35]] This could be logged on a registry to facilitate auditing and reminder systems. It would also be informative to audit antimicrobial prophylaxis for the prevention of IE against national guidelines.

It is recommended that clinicians in Northland maintain a high index of suspicion for IE and have a low threshold for performing pre-antibiotic blood cultures and TOE. A diagnosis of IE should be considered in all patients presenting with stroke or systemic emboli, particularly in the presence of fever or elevated CRP. Development of an NDHB clinical practice guideline for investigation and management of suspected IE (including indications for transfer to ADHB and inpatient dental review) and improved local access to TOE may facilitate earlier diagnosis and valve intervention. We support previous recommendations to establish a systematic New Zealand IE registry under the auspices of the Ministry of Health.[[9]] A registry would allow monitoring of IE incidence, management and outcomes at a national and district health board level and provide extensive opportunities for further auditing and quality improvement in New Zealand.

This study highlights that IE is causing alarming morbidity and mortality in Northland and consuming significant healthcare resources. Further qualitative and quantitative research into the barriers to dental care and to early diagnosis and treatment of IE in Northland may help to guide strategies for prevention and healthcare equity. Investment in equitable expansion of community water fluoridation, oral health promotion and publicly funded dental services in Northland has the potential to be cost effective for preventing IE, in addition to the marked other health benefits of improved dental health for patients.

Summary

Abstract

Aim

To explore the epidemiology, presentation, management and healthcare impact of infective endocarditis (IE) in Northland, to guide strategies for prevention and quality improvement.

Method

Health records of patients treated for IE in Northland between 2010 and 2019 were analysed retrospectively. Cases were classified using Modified Duke Diagnostic Criteria.

Results

One hundred and forty cases of IE (97 definite, 43 possible) were identified. The incidence of IE in Northland was 8.5 per 100,000-person-years. The highest-risk group were elderly Māori. There was a 44% rate of prosthetic valve endocarditis (PVE) with 27% of these patients having a history of rheumatic heart disease. Organisms causing IE included streptococcal species (43%), Staphylococcus aureus (23%) and enterococci (16%). Complications included stroke (24%), systemic embolism (38%), congestive heart failure (30%) and paravalvular abscess (14%). Median length of hospitalisation was 22 days (IQR 14–34) and 32% required valve surgery. The mortality rate at six weeks after diagnosis was 18%. An estimated total of NZ$6,560,470 was spent on direct patient care.

Conclusion

IE is causing substantial morbidity and mortality in Northland and consuming considerable healthcare resources. A high index of suspicion for IE is recommended. A high proportion of cases were caused by odontogenic organisms. Preventative investment in oral health promotion and dental care has the potential to be cost-effective.

Author Information

Johanna M Birrell: Medical Registrar, Department of Medicine, Whangārei Hospital, Whangārei. Thomas Evans: General Physician and Nephrologist, Department of Medicine, Whangārei Hospital, Whangārei. Raewyn Fisher: Cardiologist, Department of Medicine, Whangārei Hospital, Whangārei. Alan Davis: Clinical Lead for Stroke, Department of Medicine, Whangārei Hospital, Whangārei. Lucille Wilkinson: General and Obstetric Physician, Department of Medicine, Whangārei Hospital, Whangārei.

Acknowledgements

We thank all NDHB clinical and laboratory staff involved in the patients’ care. We particularly thank David Hammer for his microbiology expertise and guidance and Rowan Croft and Blair Johnson for their assistance with data extraction.

Correspondence

Johanna M Birrell, Department of Medicine, Whangārei Hospital, Maunu Rd, Private Bag 9742, Whangārei 0148.

Correspondence Email

Johanna.birrell@hbdhb.govt.nz

Competing Interests

Nil.

1)      Wong CW, Porter G,Tisch J, Young C. Outcome and prognostic factors on 57 cases of infective endocarditisin a single centre. N Z Med J. 2009;122(1304):5–62.

2)      Murdoch DR, CoreyGR, Hoen B, et al. Clinical presentation, etiology, and outcome of infective endocarditisin the 21st century: the International Collaboration on Endocarditis-ProspectiveCohort Study. Arch Intern Med. 2009;169(5):463–73.

3)      Walls G, McBride S,Raymond N, et al. Infective endocarditis in New Zealand: data from the InternationalCollaboration on Endocarditis Prospective Cohort Study. N Z Med J. 2014;127(1391):38–51.

4)      Chu J, Wilkins G,Williams M. Review of 65 cases of infective endocarditis in Dunedin Public Hospital.N Z Med J. 2004;117(1200):U1021.

5)      Yew HS, Murdoch DR.Global trends in infective endocarditis epidemiology. Curr Infect Dis Rep. 2012;14(4):367–72.

6)      Petti CA, Fowler VG,Jr. Staphylococcus aureus bacteremia and endocarditis. Cardiol Clin. 2003;21(2):219–33.

7)      Hill EE, VanderschuerenS, Verhaegen J, et al. Risk factors for infective endocarditis and outcome of patientswith Staphylococcus aureus bacteremia. Mayo Clin Proc. 2007;82(10):1165–9.

8)      Correa de Sa DD, TleyjehIM, Anavekar NS, et al. Epidemiological trends of infective endocarditis: a population-basedstudy in Olmsted County, Minnesota. Mayo Clin Proc. 2010;85(5):422–6.

9)      Kerr A, Williams M.Infective endocarditis: trends in the disease and how we study them. N Z Med J.2014;127(1391):10-2.

10)   Tleyjeh IM, Abdel-LatifA, Rahbi H, et al. A systematic review of population-based studies of infectiveendocarditis. Chest. 2007;132(3):1025-35.

11)   Yiu KH, Siu CW, LeeKL, et al. Emerging trends of community acquired infective endocarditis. Int J Cardiol.2007;121(1):119-22.

12)   Fedeli U, SchievanoE, Buonfrate D, et al. Increasing incidence and mortality of infective endocarditis:a population-based study through a record-linkage system. BMC Infect Dis. 2011;11:48.

13)   Sy RW, KritharidesL. Health care exposure and age in infective endocarditis: results of a contemporarypopulation-based profile of 1536 patients in Australia. Eur Heart J. 2010;31(15):1890-7.

14)   Lockhart PB, BrennanMT, Thornhill M, et al. Poor oral hygiene as a risk factor for infective endocarditis-relatedbacteremia. J Am Dent Assoc. 2009;140(10):1238-44.

15)   Chen SJ, Liu CJ, ChaoTF, et al. Dental scaling and risk reduction in infective endocarditis: a nationwidepopulation-based case-control study. Can J Cardiol. 2013;29(4):429-33.

16)   Ministry of Health.The New Zealand Oral Health Survey: Key Findings. Wellington: Ministry of Health;2010.

17)   Whyman RA, MahoneyEK, Stanley J. Admissions to New Zealand public hospitals for dental care: A 20-yearreview. Wellington: Ministry of Health; 2013.

18)   Gowda SS, ThomsonW, Foster Page LA, Croucher NA. Dental caries experience of children in Northland/TeTai Tokerau. N Z Dent J. 2009;105(4):116-20.

19)   Aung YM, Tin Tin S,Jelleyman T, Ameratunga S. Dental caries and previous hospitalisations among preschoolchildren: findings from a population-based study in New Zealand. N Z Med J. 2019;132(1493):44-53.

20)   Atkinson J, SalmondC, Crampton P. NZDep2018 Index of Deprivation, Interim Research Report. Wellington:University of Otago; 2019.

21)   Clark EG. Supervisedtooth brushing in Northland. Dunedin, New Zealand: University of Otago; 2017.

22)   Wauchop K, ShettyA, Bremner C. The epidemiology of acute rheumatic fever in Northland, 2012-2017.N Z Med J. 2019;132(1498):32-40.

23)   Stats NZ. 2018 CensusPlace Summaries - Northland Region. Wellington: New Zealand Government; 2018. Availablefrom: https://www.stats.govt.nz/tools/2018-census-place-summaries/northland-region.Accessed 17 May 2021.

24)   Steele JG, Walls AW.Strategies to improve the quality of oral healthcare for frail and dependent older people. Qual Health Care. 1997;6(3):165-9.

25)   Bai AD, Steinberg M, Showler A, et al. Diagnostic Accuracyof Transthoracic Echocardiography for Infective Endocarditis Findings Using TransesophagealEchocardiography as the Reference Standard: A Meta-Analysis. J Am Soc Echocardiogr.2017;30(7):639-46. e8.

26)   Durack DT, Lukes AS,Bright DK. New criteria for diagnosis of infective endocarditis: utilization ofspecific echocardiographic findings. Duke Endocarditis Service. Am J Med. 1994;96(3):200-9.

27)   Ministry of Health.Promoting Oral Health: A toolkit to assist the development, planning, implementationand evaluation of oral health promotion in New Zealand. Wellington: Ministry ofHealth; 2018. Available from: https://www.moh.govt.nz/notebook/nbbooks.nsf/0/f1f6c7e1b5c5c9f5cc257409006dbe70/$FILE/promoting-oralhealth-a-toolkit-jan08.pdf.

28)   Moore D, Poynton M,Broadbent JM, Thomson WM. The costs and benefits of water fluoridation in NZ. BMCOral Health. 2017;17(1):134.

29)   Hobbs M, Wade A, JonesP, et al. Area-level deprivation, childhood dental ambulatory sensitive hospitalizationsand community water fluoridation: evidence from New Zealand. Int J Epidemiol. 2020;49(3):908-16.

30)   NDHB. Position statement oncommunity water fluoridation. Northland: NDHB; 2016. Available from: https://www.northlanddhb.org.nz/assets/Your-Health/Northland-DHB-CWF-Position-Statement.pdf.Accessed 17 May 2021.

31)   de Silva AM, Hegde S, Akudo Nwagbara B, et al. Community-basedpopulation-level interventions for promoting child oral health. Cochrane Databaseof Systematic Reviews 2016,9. Art. No.: CD009837.

32)   Broughton JR, Maipi JTH, Person M, et al. Reducing diseaseburden and health inequalities arising from chronic disease among indigenous children:an early childhood caries intervention in Aotearoa/New Zealand. BMC Public Health.2013;13:1-8.

33)   Chen SJ, Liu CJ, Tze-Fan C, et al. Dental Scaling and RiskReduction in Infective Endocarditis: A Nationwide Population-Based Case-ControlStudy. Can J Cardiol. 2013;29(4):429-33.

34)   Infective EndocarditisProphylaxis Expert Group. Prevention of endocarditis. 2008 update from Therapeuticguidelines: Antibiotic version 13, and Therapeutic guidelines: Oral and dental version1. Melbourne: Therapeutic Guidelines Limited; 2008.

35)   Habib G, LancellottiP, Antunes MJ, et al. 2015 ESC Guidelines for the management of infective endocarditis:The Task Force for the Management of Infective Endocarditis of the European Societyof Cardiology (ESC). Endorsed by: European Association for Cardio-Thoracic Surgery(EACTS), the European Association of Nuclear Medicine (EANM). European Heart Journal.2015;36(44):3081.

For the PDF of this article,
contact nzmj@nzma.org.nz

View Article PDF

Infective endocarditis (IE) is a complex condition to diagnose and manage that continues to cause significant morbidity and mortality in New Zealand and internationally.[[1–5]] There is a trend towards an increasing proportion of IE cases caused by Staphylococcus aureus in high-income countries, while streptococcal infections continue to predominate in lower-income settings.[[2,6,7]] The International Collaboration on Endocarditis-Prospective Cohort Study (ICE-PCS) is the largest prospective cohort study on IE to date, collecting data on 2,781 patients from 58 hospitals across 25 countries (including New Zealand) between 2000 and 2005.[[2]] The ratio of streptococcal to staphylococcal species causing IE was 0.7. Other studies have demonstrated ratios of 2.5 in Pakistan and South Africa, 1.6 in South America, 1.2 in India, 0.8 in Europe and 0.3 in North America.[[5]] The higher rate of streptococcal IE in lower-income countries has been attributed to poorer dental health, higher prevalence of predisposing rheumatic heart disease (RHD), and that intravenous drug use (IVDU) and healthcare-associated risk factors for staphylococcal IE are less common.[[5,8]]

Several previous studies have explored the demographics, clinical features and management of IE in New Zealand.[[1,3,4,9]] The largest was the New Zealand component of ICE-PCS, which included 337 cases of IE.[[3]] There is a paucity of IE incidence data internationally and in New Zealand.[[5]] A systematic review of population-based studies from 1969 to 2000 across seven high-income countries reported an incidence of 1.4 to 6.2 cases per 100,000 persons.[[10]] Other studies in Hong Kong, Italy and Australia found an incidence of 2.8, 4.4 and 4.7 per 100,000 person-years respectively.[[11–13]] No previous New Zealand studies have provided population-based IE incidence data or estimated the healthcare cost of IE.

Dental disease can predispose to IE through bacteraemia from oral flora.[[14,15]] There are inequities in dental health in New Zealand, with higher rates of dental disease observed in the Māori population and among families of lower socio-economic status.[[16–19]] The Northland Region has some of the highest rates of social deprivation and dental disease in the country.[[18–20]] Dental care for adults in Northland is predominantly through private practice, other than limited emergency dentistry services for medically compromised and low-income adults in hospitals and community clinics.[[21]] There is no fluoridation of the water supply in Northland.[[21]] We hypothesise that the high burden of dental disease in Northland may be contributing to a high incidence of IE from dental organisms such as oral streptococci and enterococci.

RHD, a complication of acute rheumatic fever (ARF), is a condition associated with poverty and overcrowded living conditions that is now rarely encountered in most high-income countries. However, ARF and RHD present an ongoing challenge in Northland. A review of ARF in Northland found an incidence of 7.0 per 100,000-person-years, with 93% of cases occurring in Māori and 87% of patients living in deprived areas.[[22]] No previous studies have explored the association between RHD and IE in Northland.

This study aimed to define the epidemiology, risk factors, microbiology, presentation, management and healthcare impact of IE in Northland, to guide strategies for prevention and quality improvement.

Method

Setting

The Northland Region is defined in this study based on the 2018 census map boundaries.[[23]]Northland District Health Board (NDHB) includes five public hospitals, located in Whangārei, Bay of Islands, Kaitaia, Hokianga and Dargaville. All patients fulfilling the Modified Duke Criteria for IE were eligible for inclusion if they received treatment at any of these sites between 1 January 2010 and 31 December 2019. There is no inpatient cardiothoracic surgical service within NDHB. Northland patients requiring cardiothoracic surgical input are routinely transferred to Auckland District Health Board (ADHB).

Study design and data collection

Cases of IE were identified retrospectively using discharge coding data from NDHB hospitals (International Classification of Diseases-10 codes 133, 138, 139). Cases of all ages were included in the study if they met Modified Duke Criteria for ‘definite’ or ‘possible’ IE.[[24]] Demographic and clinical data were extracted from electronic medical records and entered on an audit tool (Microsoft® Excel (version 16.32)). New Zealand Index of Deprivation 2018 (NZDep2018) deciles were assigned according to each patient’s address at diagnosis. Territorial authority boundaries from Stats NZ were used to define the Far North, Whangārei and Kaipara districts. Nosocomial IE was defined as IE developing in a patient hospitalised for more than 48 hours prior to onset of signs or symptoms of IE.[[2]] Dental information was extracted from patients’ medical notes and the Titanium® dataset, an information management system that records clinical data from free dental examinations conducted by dental therapists and dentists.[[19]] Costing data were obtained from the NDHB webPAS® portal and included all admission costs at NDHB and ADHB for management of IE and its complications, inter-hospital transfers, cardiology clinics within 12 months after discharge and community nursing visits within 31 days of discharge.

Data analysis

As population growth occurred in Northland during the study period, population data from the approximate midpoint of the study (the mean of 2013 and 2018 census data) were used for incidence calculations. The incidence of IE was determined by dividing the number of incident cases by the total number of person-years accumulated in the study population. Incidence values and 95% confidence intervals were calculated using Microsoft Excel. Other statistical analyses were performed using IBM® SPSS Statistics (version 25.0.0.0). Continuous variables were reported as medians and interquartile ranges. Categorical variables were expressed as frequencies and percentages of the specified group. Categorical data were compared using the Chi-squared test or Fisher’s exact test. Parametric data were compared using the t-test or ANOVA test and non-parametric data using the Mann–Whitney or Kruskal–Wallis test. The Pearson correlation coefficient was used to calculate the correlation between two continuous variables.

Ethics approval

Ethics approval was sought from the New Zealand Health and Disability Ethics Committees. The study was deemed out of scope and not requiring ethics review.

Results

Incidence and demographics

The Northland population estimate used for incidence calculations was 165,384 and the Māori population was 56,807 (34%). Census data indicate 18% population growth in Northland from 2013 to 2018.[[23]] Between January 2010 and December 2019, there were 140 episodes of IE identified in 134 patients. Ninety-seven cases (69%) were classified as definite and 43 (31%) as possible IE. Table 1 describes their demographic characteristics.

The overall incidence of IE in Northland was 8.47 (95% CI 7.12–9.99)per 100,000 person-years. The incidence in males was 12.82 per 100,000 person-years,compared to 4.27 in females (p=0.04). The incidence in the New Zealand European(NZE) population was 8.30 and in Māori was 6.51 (p=0.64). Figure 1 illustrates the incidence stratified by age and ethnicity. Thepopulation group at highest risk of IE were Māori aged 80–84 years,with an incidence of 72.90 cases per 100,000 person-years compared to 18.77 amongnon-Māori in the same age group (p <0.001).Rates of IE were also significantly higher among Māori aged 25–29, 30–34,45–49 and 75–79 years.

View Table 1 & Figure 1.

One hundred and two patients (73%) lived in areas with NZDep2018 deciles between 7 and 10 (Table 1). A higher proportion of Māori than non-Māori patients lived in decile 9 and 10 areas (58% versus 39%). Figure 1 illustrates each patient’s domicile at the time of diagnosis.

Risk factors

There were 62 cases of prosthetic valve endocarditis (PVE) (44%) and 78 cases of native-valve endocarditis (56%). In 17 of the PVE cases (27%), the valve had been replaced due to RHD. Twenty-five patients with PVE (40%) had mechanical valves and 37 (60%) had bioprosthetic valves. A history of congenital valve disease was documented in 25 patients (18%) and RHD in 19 patients (14%). There were 13 cases of recurrent IE, of which six were managed for IE twice at NDHB during the study period and seven had a history of IE prior to the study period. Five cases (3%) were classified as nosocomial. Three patients (2%) had a history of injecting drug use. Further patient co-morbidity data can be found in Table 1.

Microbiology

Table 2 outlines the microbiology of IE cases according to valve type. Viridans streptococci were the most common causative organism (n=42, 30%), followed by S. aureus (n=32, 23%) and E. faecalis (n=22, 16%). The ratio of streptococcal to staphylococcal species causing IE was 1.7.

The five cases of nosocomial IE were caused by methicillin-sensitive S. aureus (MSSA) (n=1), viridans streptococci (n=1), coagulase negative staphylococci (n=1), non-HACEK group gram-negative bacilli (n=1) and polymicrobial IE (n=1).

View Table 2.

Clinical presentation

Table 3 describes clinical and investigation findings. The most frequent clinical signs were fever (n=116, 83%), heart murmur (n=92, 66%) and splinter haemorrhages (n=18, 13%). Other classical signs of IE were rare, with few patients exhibiting Janeway lesions (n=2, 1%), Roth spots (n=2, 1%) or Osler’s nodes (n=1, 1%).

In 135 cases (96%) the C-reactive protein (CRP) level was 10mg/L or higher on admission. The median CRP level on admission was 89 mg/L (IQR 52–155).

Blood cultures were positive in 129 cases (92%). Of the 11 cases with no positive blood cultures, eight had received antibiotic therapy prior to cultures being taken. The median number of blood cultures taken per patient was five (IQR 4–8), and the median number of positive blood cultures was two (IQR 1–3).

Transthoracic echocardiography (TTE) was performed in 131 patients (94%) and vegetations were identified in 48 of these (37%). Transoesophageal echocardiography (TOE) was performed in 107 patients (76%) and vegetations were identified in 77 (72%). One hundred patients (71%) underwent both TTE and TOE. Vegetations were identified on both TTE and TOE in 29 of these patients (29%) and vegetations were seen on TOE but not TTE in 48 (48%). The sensitivity of TTE compared to the gold standard of TOE in detecting valvular vegetations was 38% and the specificity was 87%. In cases of pure aortic valve IE the sensitivity of TTE compared to TOE was 45%. In cases of pure mitral valve IE the sensitivity of TTE compared to TOE was 30%. There was no significant difference in rates of TOE by ethnicity (n=25, 72% in Māori; n=81, 78% in non-Māori; p=0.490), or territorial authority (n=33, 87% in Whangārei; n=26, 76% in Far North; n=10, 91% in Kaipara District; p=0.382).

The aortic valve was involved in 79 cases (56%), mitral valve in 49 (35%), tricuspid valve in eight (6%) and pulmonary valve in four (3%). In 12 cases (9%) there were multiple valves affected. In 14 cases (10%) the affected valve(s) were unknown.

View Table 3.

Dental history

Ten cases (7%) had a recorded public dental service review in the 12 months preceding diagnosis of IE, of whom six had pre-existing prosthetic valves. Only one of the 19 cases with a history of RHD had a recorded dental review in the 12 months prior to IE diagnosis (5%). Of the 42 cases of viridans streptococcal IE, five had a recorded dental review in the prior 12 months (12%). No patients with IE caused by other streptococci or E. faecalis had a recorded dental review in the prior 12 months.

Nineteen cases (14%) underwent dental review during their hospital admission, including 13 of those with viridans streptococcal IE (31%), zero with other streptococcal IE, zero with E. faecalis IE and three with S. aureus IE. Twenty-one cases (15%) had a dental x-ray during their admission.

Complications

Complications of IE included systemic embolism (n=53, 38%), congestive heart failure (n=42, 30%), stroke (n=33, 24%), paravalvular abscess (n=19, 14%), valvular perforation (n=16, 11%), chordal vegetation or rupture (n=15, 11%), persistent bacteraemia for more than two weeks (n=10, 7%) and death within six weeks of diagnosis (n=16, 21%). There were 14 cases of prosthetic valve regurgitation (23% of all PVE cases) and 11 cases of prosthetic valve dehiscence (18% of PVE cases). An overview of complications stratified by causative organism is provided in the Table 2.

Management and outcomes

Valve surgery was performed in 45 cases (32%). Aortic valve surgery was performed in 23 patients (16%), mitral valve surgery in 26 (19%), tricuspid valve surgery in four (3%) and pulmonary valve surgery in two (1%). In 10 cases (7%), multiple valves required intervention. Further detail on types of valve surgery is provided in the Table 3. The median time from initial contact with healthcare services to valve surgery was 24 days (IQR 10–58). Valve surgery was required in 41% of S. aureus cases (n=13), 26% for viridans streptococci (n=11), 18% for E. faecalis (n=4) and 66% for coagulase negative staphylococci (n=2). Nine patients (6%) required non-cardiac surgery. Peripherally inserted central catheters were placed for 115 patients (82%) and 64 patients (46%) received outpatient intravenous antibiotic therapy.

The all-cause mortality rate at six weeks after diagnosis of IE was 18% (n=25), at six months was 22% (n=31), 12 months was 25% (n=35) and five years (or duration of follow up if less than five years) was 36% (n=51). Among patients who underwent valve surgery the six-week mortality rate was 18% (n=9), which was equivalent to the non-operative group (18%, n=16, p=0.974). In those who received ICU-level care the mortality rate was 25% (n=14), compared to 13% (n=11) in those not admitted to ICU (p=0.07). There was no significant difference in rates of ICU admission, valve surgery or death according to patients’ home territorial authority (p=0.499, 0.353 and 0.110 respectively).

Healthcare impact

Table 4 outlines the healthcare impact of IE and sites of inpatient treatment. An estimated NZ$6,560,470 was spent on direct patient care for IE during the study period. The median healthcare cost per case of IE was $34,053 (IQR $18,212–69,994).

For streptococcal IE the median cost per case was $29,362, for S. aureus was $40,420 and for E. faecalis was $30,265 (p=0.278). Length of hospital stay was directly associated with healthcare cost (correlation coefficient 0.584, p<0.000). At least one inter-hospital transfer was required for 96 of the cases (69%; range 0–6 transfers).

View Table 4.

Discussion

This study describes the burden of IE in Northland, identifying the elderly Māori population as being at greatest risk. The overall IE incidence of 8.47 per 100,000 person-years in Northland is higher than has been described in other high-income countries (reported incidence of 1.4 to 6.2 per 100,000 person-years).[[10–13]] No previous New Zealand incidence data are available for comparison.

The severity of IE cases in Northland is notable, with higher rates of systemic embolism (44% versus 17%), stroke (27% versus 14%) and death (21% versus 6%) than were observed in the New Zealand ICE-PCS cohort.[[3]] This is despite ICE-PCS only enrolling patients from large hospital centres with a cardiothoracic surgical service, which was expected to select out cases that are more complex and requiring surgical intervention.[[3]] However, the rate of valve surgery in Northland cases was similar to in the ICE-PCS cohort (34% and 33% respectively). There was a markedly longer duration from initial healthcare contact to surgery in Northland cases than in ICE-PCS (median of 24 days versus 4 days). This may be due to delays in diagnosis and inter-hospital transfer from NDHB to a cardiothoracic surgical centre, or differences in the pathology and acuity of presentation. Ethnicity and deprivation were not described for the ICE-PCS cohort. However, the rate of RHD, an indicator of underlying poverty and overcrowding, was much higher in our study (14% versus 4%) and likely reflects inequities in healthcare access and social determinants of health in Northland.

The microbiology of IE in Northland is more consistent with that seen in lower-income countries, with a high proportion of cases caused by streptococci rather than staphylococci (ratio 1.7 in Northland, versus 0.7 in the international ICE-PCS study). This pattern supports our hypothesis of dental disease in Northland contributing to the high burden of IE.

Factors contributing to the higher incidence of IE among those aged over 70 years, particularly Māori, may include a greater prevalence of cardiac valvular disease, dental decay, diabetes and immunocompromising medical conditions within this population. Changes in dental profiles with increasing retention of natural teeth may also predispose to higher rates of dental infection in later life.[[24]] Elderly people may also have difficulty accessing dental care (both for routine prevention and acute dental infections), due to disability or financial constraints. These inequities may be more pronounced among elderly Māori than non-Māori due to cultural and language barriers, remoteness and higher rates of social deprivation.[[20]]

This study also highlights the poor sensitivity of TTE (38%) compared to the gold standard of TOE in the diagnosis of IE. A systematic review and meta-analysis published in 2017 found similar results, with fundamental TTE detecting vegetations with a sensitivity of 41% (95% CI 29–55%) compared to the reference standard of TOE.[[25]] We hypothesise that reduced TOE availability in Northland may have contributed to delayed diagnosis and valve intervention. Within our cohort there was no significant difference in rates of TOE by patient ethnicity or region. However, it would be useful to compare the proportion of patients undergoing TOE and time from admission to TOE for evaluation of endocarditis between district health boards, to assess equity of access at a regional level.

It is possible that cases of IE were missed in this study if no discharge summary was completed (required for hospital coding) or the diagnosis of IE was not included during coding. Therefore, our incidence data may underestimate the true burden of IE. Another limitation of incidence calculations was the use of midpoint population estimates, as significant population growth occurred over the study period. Co-morbidity results and clinical features may be underestimated if medical history or examination was incompletely documented or not included in patients’ electronic admission or discharge records.

Only public dental data from free dental examinations were accessed. As dental care for adults in Northland is predominantly through private practice, most dental encounters are likely to have been missed (unless documented in patient notes). However, data on inpatient dental reviews at Whangārei Hospital are expected to be complete as these are routinely performed via the on-site public dental service and recorded on the Titanium database.

If patients were transferred to or from a district health board other than NDHB or ADHB (or overseas), then data from these other sites were not accessible. Assessment of longer-term outcomes such as delayed surgery, recurrence and one- and five-year mortality was not available for all patients as the time had not yet elapsed at the time of data collection.

Cases were included in the cohort if they met Modified Duke Criteria for ‘possible IE’. Some of these cases may have been incorrectly diagnosed as IE (such as episodes of bacteraemia without endocarditis). However, cases were excluded if they met Modified Duke Criteria for ‘rejected IE’ or if IE was not documented as a discharge diagnosis by the treating team.[[26]]

The calculated total healthcare cost of IE during the study period is likely a significant underestimate as it does not include non-cardiology outpatient clinic follow-up, private healthcare encounters, or the cost of long-term patient disability, interruption to employment and years of life lost. Further studies measuring the burden of disease from IE, including years of life lost and years lived with disability, would be valuable in quantifying the broader societal cost of the disease.

As a high proportion of IE cases were caused by odontogenic organisms, a review of factors contributing to dental disease and further investment in oral health promotion in Northland is recommended. The New Zealand Promoting Oral Health guideline provides a comprehensive framework for based on Ottawa Charter principles.[[27]] Community fluoridation of reticulated water supplies supplying populations over 500 has been assessed as a highly cost-effective strategy for prevention of dental caries in New Zealand.[[28]] Research indicates that populations living in deprived areas may show the greatest reduction in dental ambulatory sensitive hospitalisations as a result of community water fluoridation.[[29]] At present, no reticulated water supplies in Northland are fluoridated.[[30]] Consideration should be given to water fluoridation for dental protection in Northland, a position which is supported by NDHB.[[28,30]]

There is limited evidence that oral health education alone can improve dental hygiene.[[31]] However, a multi-component approach including oral health promotion in schools, provision of fluoride-containing toothpaste, oral health training of non-dental professionals and limitations on marketing and sale of high-sugar food and beverages may be beneficial.[[31]] Further study is currently underway to develop and evaluate interventions to reduce oral health disparities for Māori.[[32]]

Previous research suggests that improvement in oral hygiene by regular dental scaling is associated with a significant decrease in risk of IE.[[33]] Access to affordable dental care is particularly important for high-risk groups such as elderly Māori and patients with prosthetic valves, RHD, congenital valve disease or previous IE. International guidelines recommend that high-risk patients undergo dental follow-up twice a year.[[34,35]] This could be logged on a registry to facilitate auditing and reminder systems. It would also be informative to audit antimicrobial prophylaxis for the prevention of IE against national guidelines.

It is recommended that clinicians in Northland maintain a high index of suspicion for IE and have a low threshold for performing pre-antibiotic blood cultures and TOE. A diagnosis of IE should be considered in all patients presenting with stroke or systemic emboli, particularly in the presence of fever or elevated CRP. Development of an NDHB clinical practice guideline for investigation and management of suspected IE (including indications for transfer to ADHB and inpatient dental review) and improved local access to TOE may facilitate earlier diagnosis and valve intervention. We support previous recommendations to establish a systematic New Zealand IE registry under the auspices of the Ministry of Health.[[9]] A registry would allow monitoring of IE incidence, management and outcomes at a national and district health board level and provide extensive opportunities for further auditing and quality improvement in New Zealand.

This study highlights that IE is causing alarming morbidity and mortality in Northland and consuming significant healthcare resources. Further qualitative and quantitative research into the barriers to dental care and to early diagnosis and treatment of IE in Northland may help to guide strategies for prevention and healthcare equity. Investment in equitable expansion of community water fluoridation, oral health promotion and publicly funded dental services in Northland has the potential to be cost effective for preventing IE, in addition to the marked other health benefits of improved dental health for patients.

Summary

Abstract

Aim

To explore the epidemiology, presentation, management and healthcare impact of infective endocarditis (IE) in Northland, to guide strategies for prevention and quality improvement.

Method

Health records of patients treated for IE in Northland between 2010 and 2019 were analysed retrospectively. Cases were classified using Modified Duke Diagnostic Criteria.

Results

One hundred and forty cases of IE (97 definite, 43 possible) were identified. The incidence of IE in Northland was 8.5 per 100,000-person-years. The highest-risk group were elderly Māori. There was a 44% rate of prosthetic valve endocarditis (PVE) with 27% of these patients having a history of rheumatic heart disease. Organisms causing IE included streptococcal species (43%), Staphylococcus aureus (23%) and enterococci (16%). Complications included stroke (24%), systemic embolism (38%), congestive heart failure (30%) and paravalvular abscess (14%). Median length of hospitalisation was 22 days (IQR 14–34) and 32% required valve surgery. The mortality rate at six weeks after diagnosis was 18%. An estimated total of NZ$6,560,470 was spent on direct patient care.

Conclusion

IE is causing substantial morbidity and mortality in Northland and consuming considerable healthcare resources. A high index of suspicion for IE is recommended. A high proportion of cases were caused by odontogenic organisms. Preventative investment in oral health promotion and dental care has the potential to be cost-effective.

Author Information

Johanna M Birrell: Medical Registrar, Department of Medicine, Whangārei Hospital, Whangārei. Thomas Evans: General Physician and Nephrologist, Department of Medicine, Whangārei Hospital, Whangārei. Raewyn Fisher: Cardiologist, Department of Medicine, Whangārei Hospital, Whangārei. Alan Davis: Clinical Lead for Stroke, Department of Medicine, Whangārei Hospital, Whangārei. Lucille Wilkinson: General and Obstetric Physician, Department of Medicine, Whangārei Hospital, Whangārei.

Acknowledgements

We thank all NDHB clinical and laboratory staff involved in the patients’ care. We particularly thank David Hammer for his microbiology expertise and guidance and Rowan Croft and Blair Johnson for their assistance with data extraction.

Correspondence

Johanna M Birrell, Department of Medicine, Whangārei Hospital, Maunu Rd, Private Bag 9742, Whangārei 0148.

Correspondence Email

Johanna.birrell@hbdhb.govt.nz

Competing Interests

Nil.

1)      Wong CW, Porter G,Tisch J, Young C. Outcome and prognostic factors on 57 cases of infective endocarditisin a single centre. N Z Med J. 2009;122(1304):5–62.

2)      Murdoch DR, CoreyGR, Hoen B, et al. Clinical presentation, etiology, and outcome of infective endocarditisin the 21st century: the International Collaboration on Endocarditis-ProspectiveCohort Study. Arch Intern Med. 2009;169(5):463–73.

3)      Walls G, McBride S,Raymond N, et al. Infective endocarditis in New Zealand: data from the InternationalCollaboration on Endocarditis Prospective Cohort Study. N Z Med J. 2014;127(1391):38–51.

4)      Chu J, Wilkins G,Williams M. Review of 65 cases of infective endocarditis in Dunedin Public Hospital.N Z Med J. 2004;117(1200):U1021.

5)      Yew HS, Murdoch DR.Global trends in infective endocarditis epidemiology. Curr Infect Dis Rep. 2012;14(4):367–72.

6)      Petti CA, Fowler VG,Jr. Staphylococcus aureus bacteremia and endocarditis. Cardiol Clin. 2003;21(2):219–33.

7)      Hill EE, VanderschuerenS, Verhaegen J, et al. Risk factors for infective endocarditis and outcome of patientswith Staphylococcus aureus bacteremia. Mayo Clin Proc. 2007;82(10):1165–9.

8)      Correa de Sa DD, TleyjehIM, Anavekar NS, et al. Epidemiological trends of infective endocarditis: a population-basedstudy in Olmsted County, Minnesota. Mayo Clin Proc. 2010;85(5):422–6.

9)      Kerr A, Williams M.Infective endocarditis: trends in the disease and how we study them. N Z Med J.2014;127(1391):10-2.

10)   Tleyjeh IM, Abdel-LatifA, Rahbi H, et al. A systematic review of population-based studies of infectiveendocarditis. Chest. 2007;132(3):1025-35.

11)   Yiu KH, Siu CW, LeeKL, et al. Emerging trends of community acquired infective endocarditis. Int J Cardiol.2007;121(1):119-22.

12)   Fedeli U, SchievanoE, Buonfrate D, et al. Increasing incidence and mortality of infective endocarditis:a population-based study through a record-linkage system. BMC Infect Dis. 2011;11:48.

13)   Sy RW, KritharidesL. Health care exposure and age in infective endocarditis: results of a contemporarypopulation-based profile of 1536 patients in Australia. Eur Heart J. 2010;31(15):1890-7.

14)   Lockhart PB, BrennanMT, Thornhill M, et al. Poor oral hygiene as a risk factor for infective endocarditis-relatedbacteremia. J Am Dent Assoc. 2009;140(10):1238-44.

15)   Chen SJ, Liu CJ, ChaoTF, et al. Dental scaling and risk reduction in infective endocarditis: a nationwidepopulation-based case-control study. Can J Cardiol. 2013;29(4):429-33.

16)   Ministry of Health.The New Zealand Oral Health Survey: Key Findings. Wellington: Ministry of Health;2010.

17)   Whyman RA, MahoneyEK, Stanley J. Admissions to New Zealand public hospitals for dental care: A 20-yearreview. Wellington: Ministry of Health; 2013.

18)   Gowda SS, ThomsonW, Foster Page LA, Croucher NA. Dental caries experience of children in Northland/TeTai Tokerau. N Z Dent J. 2009;105(4):116-20.

19)   Aung YM, Tin Tin S,Jelleyman T, Ameratunga S. Dental caries and previous hospitalisations among preschoolchildren: findings from a population-based study in New Zealand. N Z Med J. 2019;132(1493):44-53.

20)   Atkinson J, SalmondC, Crampton P. NZDep2018 Index of Deprivation, Interim Research Report. Wellington:University of Otago; 2019.

21)   Clark EG. Supervisedtooth brushing in Northland. Dunedin, New Zealand: University of Otago; 2017.

22)   Wauchop K, ShettyA, Bremner C. The epidemiology of acute rheumatic fever in Northland, 2012-2017.N Z Med J. 2019;132(1498):32-40.

23)   Stats NZ. 2018 CensusPlace Summaries - Northland Region. Wellington: New Zealand Government; 2018. Availablefrom: https://www.stats.govt.nz/tools/2018-census-place-summaries/northland-region.Accessed 17 May 2021.

24)   Steele JG, Walls AW.Strategies to improve the quality of oral healthcare for frail and dependent older people. Qual Health Care. 1997;6(3):165-9.

25)   Bai AD, Steinberg M, Showler A, et al. Diagnostic Accuracyof Transthoracic Echocardiography for Infective Endocarditis Findings Using TransesophagealEchocardiography as the Reference Standard: A Meta-Analysis. J Am Soc Echocardiogr.2017;30(7):639-46. e8.

26)   Durack DT, Lukes AS,Bright DK. New criteria for diagnosis of infective endocarditis: utilization ofspecific echocardiographic findings. Duke Endocarditis Service. Am J Med. 1994;96(3):200-9.

27)   Ministry of Health.Promoting Oral Health: A toolkit to assist the development, planning, implementationand evaluation of oral health promotion in New Zealand. Wellington: Ministry ofHealth; 2018. Available from: https://www.moh.govt.nz/notebook/nbbooks.nsf/0/f1f6c7e1b5c5c9f5cc257409006dbe70/$FILE/promoting-oralhealth-a-toolkit-jan08.pdf.

28)   Moore D, Poynton M,Broadbent JM, Thomson WM. The costs and benefits of water fluoridation in NZ. BMCOral Health. 2017;17(1):134.

29)   Hobbs M, Wade A, JonesP, et al. Area-level deprivation, childhood dental ambulatory sensitive hospitalizationsand community water fluoridation: evidence from New Zealand. Int J Epidemiol. 2020;49(3):908-16.

30)   NDHB. Position statement oncommunity water fluoridation. Northland: NDHB; 2016. Available from: https://www.northlanddhb.org.nz/assets/Your-Health/Northland-DHB-CWF-Position-Statement.pdf.Accessed 17 May 2021.

31)   de Silva AM, Hegde S, Akudo Nwagbara B, et al. Community-basedpopulation-level interventions for promoting child oral health. Cochrane Databaseof Systematic Reviews 2016,9. Art. No.: CD009837.

32)   Broughton JR, Maipi JTH, Person M, et al. Reducing diseaseburden and health inequalities arising from chronic disease among indigenous children:an early childhood caries intervention in Aotearoa/New Zealand. BMC Public Health.2013;13:1-8.

33)   Chen SJ, Liu CJ, Tze-Fan C, et al. Dental Scaling and RiskReduction in Infective Endocarditis: A Nationwide Population-Based Case-ControlStudy. Can J Cardiol. 2013;29(4):429-33.

34)   Infective EndocarditisProphylaxis Expert Group. Prevention of endocarditis. 2008 update from Therapeuticguidelines: Antibiotic version 13, and Therapeutic guidelines: Oral and dental version1. Melbourne: Therapeutic Guidelines Limited; 2008.

35)   Habib G, LancellottiP, Antunes MJ, et al. 2015 ESC Guidelines for the management of infective endocarditis:The Task Force for the Management of Infective Endocarditis of the European Societyof Cardiology (ESC). Endorsed by: European Association for Cardio-Thoracic Surgery(EACTS), the European Association of Nuclear Medicine (EANM). European Heart Journal.2015;36(44):3081.

For the PDF of this article,
contact nzmj@nzma.org.nz

View Article PDF

Infective endocarditis (IE) is a complex condition to diagnose and manage that continues to cause significant morbidity and mortality in New Zealand and internationally.[[1–5]] There is a trend towards an increasing proportion of IE cases caused by Staphylococcus aureus in high-income countries, while streptococcal infections continue to predominate in lower-income settings.[[2,6,7]] The International Collaboration on Endocarditis-Prospective Cohort Study (ICE-PCS) is the largest prospective cohort study on IE to date, collecting data on 2,781 patients from 58 hospitals across 25 countries (including New Zealand) between 2000 and 2005.[[2]] The ratio of streptococcal to staphylococcal species causing IE was 0.7. Other studies have demonstrated ratios of 2.5 in Pakistan and South Africa, 1.6 in South America, 1.2 in India, 0.8 in Europe and 0.3 in North America.[[5]] The higher rate of streptococcal IE in lower-income countries has been attributed to poorer dental health, higher prevalence of predisposing rheumatic heart disease (RHD), and that intravenous drug use (IVDU) and healthcare-associated risk factors for staphylococcal IE are less common.[[5,8]]

Several previous studies have explored the demographics, clinical features and management of IE in New Zealand.[[1,3,4,9]] The largest was the New Zealand component of ICE-PCS, which included 337 cases of IE.[[3]] There is a paucity of IE incidence data internationally and in New Zealand.[[5]] A systematic review of population-based studies from 1969 to 2000 across seven high-income countries reported an incidence of 1.4 to 6.2 cases per 100,000 persons.[[10]] Other studies in Hong Kong, Italy and Australia found an incidence of 2.8, 4.4 and 4.7 per 100,000 person-years respectively.[[11–13]] No previous New Zealand studies have provided population-based IE incidence data or estimated the healthcare cost of IE.

Dental disease can predispose to IE through bacteraemia from oral flora.[[14,15]] There are inequities in dental health in New Zealand, with higher rates of dental disease observed in the Māori population and among families of lower socio-economic status.[[16–19]] The Northland Region has some of the highest rates of social deprivation and dental disease in the country.[[18–20]] Dental care for adults in Northland is predominantly through private practice, other than limited emergency dentistry services for medically compromised and low-income adults in hospitals and community clinics.[[21]] There is no fluoridation of the water supply in Northland.[[21]] We hypothesise that the high burden of dental disease in Northland may be contributing to a high incidence of IE from dental organisms such as oral streptococci and enterococci.

RHD, a complication of acute rheumatic fever (ARF), is a condition associated with poverty and overcrowded living conditions that is now rarely encountered in most high-income countries. However, ARF and RHD present an ongoing challenge in Northland. A review of ARF in Northland found an incidence of 7.0 per 100,000-person-years, with 93% of cases occurring in Māori and 87% of patients living in deprived areas.[[22]] No previous studies have explored the association between RHD and IE in Northland.

This study aimed to define the epidemiology, risk factors, microbiology, presentation, management and healthcare impact of IE in Northland, to guide strategies for prevention and quality improvement.

Method

Setting

The Northland Region is defined in this study based on the 2018 census map boundaries.[[23]]Northland District Health Board (NDHB) includes five public hospitals, located in Whangārei, Bay of Islands, Kaitaia, Hokianga and Dargaville. All patients fulfilling the Modified Duke Criteria for IE were eligible for inclusion if they received treatment at any of these sites between 1 January 2010 and 31 December 2019. There is no inpatient cardiothoracic surgical service within NDHB. Northland patients requiring cardiothoracic surgical input are routinely transferred to Auckland District Health Board (ADHB).

Study design and data collection

Cases of IE were identified retrospectively using discharge coding data from NDHB hospitals (International Classification of Diseases-10 codes 133, 138, 139). Cases of all ages were included in the study if they met Modified Duke Criteria for ‘definite’ or ‘possible’ IE.[[24]] Demographic and clinical data were extracted from electronic medical records and entered on an audit tool (Microsoft® Excel (version 16.32)). New Zealand Index of Deprivation 2018 (NZDep2018) deciles were assigned according to each patient’s address at diagnosis. Territorial authority boundaries from Stats NZ were used to define the Far North, Whangārei and Kaipara districts. Nosocomial IE was defined as IE developing in a patient hospitalised for more than 48 hours prior to onset of signs or symptoms of IE.[[2]] Dental information was extracted from patients’ medical notes and the Titanium® dataset, an information management system that records clinical data from free dental examinations conducted by dental therapists and dentists.[[19]] Costing data were obtained from the NDHB webPAS® portal and included all admission costs at NDHB and ADHB for management of IE and its complications, inter-hospital transfers, cardiology clinics within 12 months after discharge and community nursing visits within 31 days of discharge.

Data analysis

As population growth occurred in Northland during the study period, population data from the approximate midpoint of the study (the mean of 2013 and 2018 census data) were used for incidence calculations. The incidence of IE was determined by dividing the number of incident cases by the total number of person-years accumulated in the study population. Incidence values and 95% confidence intervals were calculated using Microsoft Excel. Other statistical analyses were performed using IBM® SPSS Statistics (version 25.0.0.0). Continuous variables were reported as medians and interquartile ranges. Categorical variables were expressed as frequencies and percentages of the specified group. Categorical data were compared using the Chi-squared test or Fisher’s exact test. Parametric data were compared using the t-test or ANOVA test and non-parametric data using the Mann–Whitney or Kruskal–Wallis test. The Pearson correlation coefficient was used to calculate the correlation between two continuous variables.

Ethics approval

Ethics approval was sought from the New Zealand Health and Disability Ethics Committees. The study was deemed out of scope and not requiring ethics review.

Results

Incidence and demographics

The Northland population estimate used for incidence calculations was 165,384 and the Māori population was 56,807 (34%). Census data indicate 18% population growth in Northland from 2013 to 2018.[[23]] Between January 2010 and December 2019, there were 140 episodes of IE identified in 134 patients. Ninety-seven cases (69%) were classified as definite and 43 (31%) as possible IE. Table 1 describes their demographic characteristics.

The overall incidence of IE in Northland was 8.47 (95% CI 7.12–9.99)per 100,000 person-years. The incidence in males was 12.82 per 100,000 person-years,compared to 4.27 in females (p=0.04). The incidence in the New Zealand European(NZE) population was 8.30 and in Māori was 6.51 (p=0.64). Figure 1 illustrates the incidence stratified by age and ethnicity. Thepopulation group at highest risk of IE were Māori aged 80–84 years,with an incidence of 72.90 cases per 100,000 person-years compared to 18.77 amongnon-Māori in the same age group (p <0.001).Rates of IE were also significantly higher among Māori aged 25–29, 30–34,45–49 and 75–79 years.

View Table 1 & Figure 1.

One hundred and two patients (73%) lived in areas with NZDep2018 deciles between 7 and 10 (Table 1). A higher proportion of Māori than non-Māori patients lived in decile 9 and 10 areas (58% versus 39%). Figure 1 illustrates each patient’s domicile at the time of diagnosis.

Risk factors

There were 62 cases of prosthetic valve endocarditis (PVE) (44%) and 78 cases of native-valve endocarditis (56%). In 17 of the PVE cases (27%), the valve had been replaced due to RHD. Twenty-five patients with PVE (40%) had mechanical valves and 37 (60%) had bioprosthetic valves. A history of congenital valve disease was documented in 25 patients (18%) and RHD in 19 patients (14%). There were 13 cases of recurrent IE, of which six were managed for IE twice at NDHB during the study period and seven had a history of IE prior to the study period. Five cases (3%) were classified as nosocomial. Three patients (2%) had a history of injecting drug use. Further patient co-morbidity data can be found in Table 1.

Microbiology

Table 2 outlines the microbiology of IE cases according to valve type. Viridans streptococci were the most common causative organism (n=42, 30%), followed by S. aureus (n=32, 23%) and E. faecalis (n=22, 16%). The ratio of streptococcal to staphylococcal species causing IE was 1.7.

The five cases of nosocomial IE were caused by methicillin-sensitive S. aureus (MSSA) (n=1), viridans streptococci (n=1), coagulase negative staphylococci (n=1), non-HACEK group gram-negative bacilli (n=1) and polymicrobial IE (n=1).

View Table 2.

Clinical presentation

Table 3 describes clinical and investigation findings. The most frequent clinical signs were fever (n=116, 83%), heart murmur (n=92, 66%) and splinter haemorrhages (n=18, 13%). Other classical signs of IE were rare, with few patients exhibiting Janeway lesions (n=2, 1%), Roth spots (n=2, 1%) or Osler’s nodes (n=1, 1%).

In 135 cases (96%) the C-reactive protein (CRP) level was 10mg/L or higher on admission. The median CRP level on admission was 89 mg/L (IQR 52–155).

Blood cultures were positive in 129 cases (92%). Of the 11 cases with no positive blood cultures, eight had received antibiotic therapy prior to cultures being taken. The median number of blood cultures taken per patient was five (IQR 4–8), and the median number of positive blood cultures was two (IQR 1–3).

Transthoracic echocardiography (TTE) was performed in 131 patients (94%) and vegetations were identified in 48 of these (37%). Transoesophageal echocardiography (TOE) was performed in 107 patients (76%) and vegetations were identified in 77 (72%). One hundred patients (71%) underwent both TTE and TOE. Vegetations were identified on both TTE and TOE in 29 of these patients (29%) and vegetations were seen on TOE but not TTE in 48 (48%). The sensitivity of TTE compared to the gold standard of TOE in detecting valvular vegetations was 38% and the specificity was 87%. In cases of pure aortic valve IE the sensitivity of TTE compared to TOE was 45%. In cases of pure mitral valve IE the sensitivity of TTE compared to TOE was 30%. There was no significant difference in rates of TOE by ethnicity (n=25, 72% in Māori; n=81, 78% in non-Māori; p=0.490), or territorial authority (n=33, 87% in Whangārei; n=26, 76% in Far North; n=10, 91% in Kaipara District; p=0.382).

The aortic valve was involved in 79 cases (56%), mitral valve in 49 (35%), tricuspid valve in eight (6%) and pulmonary valve in four (3%). In 12 cases (9%) there were multiple valves affected. In 14 cases (10%) the affected valve(s) were unknown.

View Table 3.

Dental history

Ten cases (7%) had a recorded public dental service review in the 12 months preceding diagnosis of IE, of whom six had pre-existing prosthetic valves. Only one of the 19 cases with a history of RHD had a recorded dental review in the 12 months prior to IE diagnosis (5%). Of the 42 cases of viridans streptococcal IE, five had a recorded dental review in the prior 12 months (12%). No patients with IE caused by other streptococci or E. faecalis had a recorded dental review in the prior 12 months.

Nineteen cases (14%) underwent dental review during their hospital admission, including 13 of those with viridans streptococcal IE (31%), zero with other streptococcal IE, zero with E. faecalis IE and three with S. aureus IE. Twenty-one cases (15%) had a dental x-ray during their admission.

Complications

Complications of IE included systemic embolism (n=53, 38%), congestive heart failure (n=42, 30%), stroke (n=33, 24%), paravalvular abscess (n=19, 14%), valvular perforation (n=16, 11%), chordal vegetation or rupture (n=15, 11%), persistent bacteraemia for more than two weeks (n=10, 7%) and death within six weeks of diagnosis (n=16, 21%). There were 14 cases of prosthetic valve regurgitation (23% of all PVE cases) and 11 cases of prosthetic valve dehiscence (18% of PVE cases). An overview of complications stratified by causative organism is provided in the Table 2.

Management and outcomes

Valve surgery was performed in 45 cases (32%). Aortic valve surgery was performed in 23 patients (16%), mitral valve surgery in 26 (19%), tricuspid valve surgery in four (3%) and pulmonary valve surgery in two (1%). In 10 cases (7%), multiple valves required intervention. Further detail on types of valve surgery is provided in the Table 3. The median time from initial contact with healthcare services to valve surgery was 24 days (IQR 10–58). Valve surgery was required in 41% of S. aureus cases (n=13), 26% for viridans streptococci (n=11), 18% for E. faecalis (n=4) and 66% for coagulase negative staphylococci (n=2). Nine patients (6%) required non-cardiac surgery. Peripherally inserted central catheters were placed for 115 patients (82%) and 64 patients (46%) received outpatient intravenous antibiotic therapy.

The all-cause mortality rate at six weeks after diagnosis of IE was 18% (n=25), at six months was 22% (n=31), 12 months was 25% (n=35) and five years (or duration of follow up if less than five years) was 36% (n=51). Among patients who underwent valve surgery the six-week mortality rate was 18% (n=9), which was equivalent to the non-operative group (18%, n=16, p=0.974). In those who received ICU-level care the mortality rate was 25% (n=14), compared to 13% (n=11) in those not admitted to ICU (p=0.07). There was no significant difference in rates of ICU admission, valve surgery or death according to patients’ home territorial authority (p=0.499, 0.353 and 0.110 respectively).

Healthcare impact

Table 4 outlines the healthcare impact of IE and sites of inpatient treatment. An estimated NZ$6,560,470 was spent on direct patient care for IE during the study period. The median healthcare cost per case of IE was $34,053 (IQR $18,212–69,994).

For streptococcal IE the median cost per case was $29,362, for S. aureus was $40,420 and for E. faecalis was $30,265 (p=0.278). Length of hospital stay was directly associated with healthcare cost (correlation coefficient 0.584, p<0.000). At least one inter-hospital transfer was required for 96 of the cases (69%; range 0–6 transfers).

View Table 4.

Discussion

This study describes the burden of IE in Northland, identifying the elderly Māori population as being at greatest risk. The overall IE incidence of 8.47 per 100,000 person-years in Northland is higher than has been described in other high-income countries (reported incidence of 1.4 to 6.2 per 100,000 person-years).[[10–13]] No previous New Zealand incidence data are available for comparison.

The severity of IE cases in Northland is notable, with higher rates of systemic embolism (44% versus 17%), stroke (27% versus 14%) and death (21% versus 6%) than were observed in the New Zealand ICE-PCS cohort.[[3]] This is despite ICE-PCS only enrolling patients from large hospital centres with a cardiothoracic surgical service, which was expected to select out cases that are more complex and requiring surgical intervention.[[3]] However, the rate of valve surgery in Northland cases was similar to in the ICE-PCS cohort (34% and 33% respectively). There was a markedly longer duration from initial healthcare contact to surgery in Northland cases than in ICE-PCS (median of 24 days versus 4 days). This may be due to delays in diagnosis and inter-hospital transfer from NDHB to a cardiothoracic surgical centre, or differences in the pathology and acuity of presentation. Ethnicity and deprivation were not described for the ICE-PCS cohort. However, the rate of RHD, an indicator of underlying poverty and overcrowding, was much higher in our study (14% versus 4%) and likely reflects inequities in healthcare access and social determinants of health in Northland.

The microbiology of IE in Northland is more consistent with that seen in lower-income countries, with a high proportion of cases caused by streptococci rather than staphylococci (ratio 1.7 in Northland, versus 0.7 in the international ICE-PCS study). This pattern supports our hypothesis of dental disease in Northland contributing to the high burden of IE.

Factors contributing to the higher incidence of IE among those aged over 70 years, particularly Māori, may include a greater prevalence of cardiac valvular disease, dental decay, diabetes and immunocompromising medical conditions within this population. Changes in dental profiles with increasing retention of natural teeth may also predispose to higher rates of dental infection in later life.[[24]] Elderly people may also have difficulty accessing dental care (both for routine prevention and acute dental infections), due to disability or financial constraints. These inequities may be more pronounced among elderly Māori than non-Māori due to cultural and language barriers, remoteness and higher rates of social deprivation.[[20]]

This study also highlights the poor sensitivity of TTE (38%) compared to the gold standard of TOE in the diagnosis of IE. A systematic review and meta-analysis published in 2017 found similar results, with fundamental TTE detecting vegetations with a sensitivity of 41% (95% CI 29–55%) compared to the reference standard of TOE.[[25]] We hypothesise that reduced TOE availability in Northland may have contributed to delayed diagnosis and valve intervention. Within our cohort there was no significant difference in rates of TOE by patient ethnicity or region. However, it would be useful to compare the proportion of patients undergoing TOE and time from admission to TOE for evaluation of endocarditis between district health boards, to assess equity of access at a regional level.

It is possible that cases of IE were missed in this study if no discharge summary was completed (required for hospital coding) or the diagnosis of IE was not included during coding. Therefore, our incidence data may underestimate the true burden of IE. Another limitation of incidence calculations was the use of midpoint population estimates, as significant population growth occurred over the study period. Co-morbidity results and clinical features may be underestimated if medical history or examination was incompletely documented or not included in patients’ electronic admission or discharge records.

Only public dental data from free dental examinations were accessed. As dental care for adults in Northland is predominantly through private practice, most dental encounters are likely to have been missed (unless documented in patient notes). However, data on inpatient dental reviews at Whangārei Hospital are expected to be complete as these are routinely performed via the on-site public dental service and recorded on the Titanium database.

If patients were transferred to or from a district health board other than NDHB or ADHB (or overseas), then data from these other sites were not accessible. Assessment of longer-term outcomes such as delayed surgery, recurrence and one- and five-year mortality was not available for all patients as the time had not yet elapsed at the time of data collection.

Cases were included in the cohort if they met Modified Duke Criteria for ‘possible IE’. Some of these cases may have been incorrectly diagnosed as IE (such as episodes of bacteraemia without endocarditis). However, cases were excluded if they met Modified Duke Criteria for ‘rejected IE’ or if IE was not documented as a discharge diagnosis by the treating team.[[26]]

The calculated total healthcare cost of IE during the study period is likely a significant underestimate as it does not include non-cardiology outpatient clinic follow-up, private healthcare encounters, or the cost of long-term patient disability, interruption to employment and years of life lost. Further studies measuring the burden of disease from IE, including years of life lost and years lived with disability, would be valuable in quantifying the broader societal cost of the disease.

As a high proportion of IE cases were caused by odontogenic organisms, a review of factors contributing to dental disease and further investment in oral health promotion in Northland is recommended. The New Zealand Promoting Oral Health guideline provides a comprehensive framework for based on Ottawa Charter principles.[[27]] Community fluoridation of reticulated water supplies supplying populations over 500 has been assessed as a highly cost-effective strategy for prevention of dental caries in New Zealand.[[28]] Research indicates that populations living in deprived areas may show the greatest reduction in dental ambulatory sensitive hospitalisations as a result of community water fluoridation.[[29]] At present, no reticulated water supplies in Northland are fluoridated.[[30]] Consideration should be given to water fluoridation for dental protection in Northland, a position which is supported by NDHB.[[28,30]]

There is limited evidence that oral health education alone can improve dental hygiene.[[31]] However, a multi-component approach including oral health promotion in schools, provision of fluoride-containing toothpaste, oral health training of non-dental professionals and limitations on marketing and sale of high-sugar food and beverages may be beneficial.[[31]] Further study is currently underway to develop and evaluate interventions to reduce oral health disparities for Māori.[[32]]

Previous research suggests that improvement in oral hygiene by regular dental scaling is associated with a significant decrease in risk of IE.[[33]] Access to affordable dental care is particularly important for high-risk groups such as elderly Māori and patients with prosthetic valves, RHD, congenital valve disease or previous IE. International guidelines recommend that high-risk patients undergo dental follow-up twice a year.[[34,35]] This could be logged on a registry to facilitate auditing and reminder systems. It would also be informative to audit antimicrobial prophylaxis for the prevention of IE against national guidelines.

It is recommended that clinicians in Northland maintain a high index of suspicion for IE and have a low threshold for performing pre-antibiotic blood cultures and TOE. A diagnosis of IE should be considered in all patients presenting with stroke or systemic emboli, particularly in the presence of fever or elevated CRP. Development of an NDHB clinical practice guideline for investigation and management of suspected IE (including indications for transfer to ADHB and inpatient dental review) and improved local access to TOE may facilitate earlier diagnosis and valve intervention. We support previous recommendations to establish a systematic New Zealand IE registry under the auspices of the Ministry of Health.[[9]] A registry would allow monitoring of IE incidence, management and outcomes at a national and district health board level and provide extensive opportunities for further auditing and quality improvement in New Zealand.

This study highlights that IE is causing alarming morbidity and mortality in Northland and consuming significant healthcare resources. Further qualitative and quantitative research into the barriers to dental care and to early diagnosis and treatment of IE in Northland may help to guide strategies for prevention and healthcare equity. Investment in equitable expansion of community water fluoridation, oral health promotion and publicly funded dental services in Northland has the potential to be cost effective for preventing IE, in addition to the marked other health benefits of improved dental health for patients.

Summary

Abstract

Aim

To explore the epidemiology, presentation, management and healthcare impact of infective endocarditis (IE) in Northland, to guide strategies for prevention and quality improvement.

Method

Health records of patients treated for IE in Northland between 2010 and 2019 were analysed retrospectively. Cases were classified using Modified Duke Diagnostic Criteria.

Results

One hundred and forty cases of IE (97 definite, 43 possible) were identified. The incidence of IE in Northland was 8.5 per 100,000-person-years. The highest-risk group were elderly Māori. There was a 44% rate of prosthetic valve endocarditis (PVE) with 27% of these patients having a history of rheumatic heart disease. Organisms causing IE included streptococcal species (43%), Staphylococcus aureus (23%) and enterococci (16%). Complications included stroke (24%), systemic embolism (38%), congestive heart failure (30%) and paravalvular abscess (14%). Median length of hospitalisation was 22 days (IQR 14–34) and 32% required valve surgery. The mortality rate at six weeks after diagnosis was 18%. An estimated total of NZ$6,560,470 was spent on direct patient care.

Conclusion

IE is causing substantial morbidity and mortality in Northland and consuming considerable healthcare resources. A high index of suspicion for IE is recommended. A high proportion of cases were caused by odontogenic organisms. Preventative investment in oral health promotion and dental care has the potential to be cost-effective.

Author Information

Johanna M Birrell: Medical Registrar, Department of Medicine, Whangārei Hospital, Whangārei. Thomas Evans: General Physician and Nephrologist, Department of Medicine, Whangārei Hospital, Whangārei. Raewyn Fisher: Cardiologist, Department of Medicine, Whangārei Hospital, Whangārei. Alan Davis: Clinical Lead for Stroke, Department of Medicine, Whangārei Hospital, Whangārei. Lucille Wilkinson: General and Obstetric Physician, Department of Medicine, Whangārei Hospital, Whangārei.

Acknowledgements

We thank all NDHB clinical and laboratory staff involved in the patients’ care. We particularly thank David Hammer for his microbiology expertise and guidance and Rowan Croft and Blair Johnson for their assistance with data extraction.

Correspondence

Johanna M Birrell, Department of Medicine, Whangārei Hospital, Maunu Rd, Private Bag 9742, Whangārei 0148.

Correspondence Email

Johanna.birrell@hbdhb.govt.nz

Competing Interests

Nil.

1)      Wong CW, Porter G,Tisch J, Young C. Outcome and prognostic factors on 57 cases of infective endocarditisin a single centre. N Z Med J. 2009;122(1304):5–62.

2)      Murdoch DR, CoreyGR, Hoen B, et al. Clinical presentation, etiology, and outcome of infective endocarditisin the 21st century: the International Collaboration on Endocarditis-ProspectiveCohort Study. Arch Intern Med. 2009;169(5):463–73.

3)      Walls G, McBride S,Raymond N, et al. Infective endocarditis in New Zealand: data from the InternationalCollaboration on Endocarditis Prospective Cohort Study. N Z Med J. 2014;127(1391):38–51.

4)      Chu J, Wilkins G,Williams M. Review of 65 cases of infective endocarditis in Dunedin Public Hospital.N Z Med J. 2004;117(1200):U1021.

5)      Yew HS, Murdoch DR.Global trends in infective endocarditis epidemiology. Curr Infect Dis Rep. 2012;14(4):367–72.

6)      Petti CA, Fowler VG,Jr. Staphylococcus aureus bacteremia and endocarditis. Cardiol Clin. 2003;21(2):219–33.

7)      Hill EE, VanderschuerenS, Verhaegen J, et al. Risk factors for infective endocarditis and outcome of patientswith Staphylococcus aureus bacteremia. Mayo Clin Proc. 2007;82(10):1165–9.

8)      Correa de Sa DD, TleyjehIM, Anavekar NS, et al. Epidemiological trends of infective endocarditis: a population-basedstudy in Olmsted County, Minnesota. Mayo Clin Proc. 2010;85(5):422–6.

9)      Kerr A, Williams M.Infective endocarditis: trends in the disease and how we study them. N Z Med J.2014;127(1391):10-2.

10)   Tleyjeh IM, Abdel-LatifA, Rahbi H, et al. A systematic review of population-based studies of infectiveendocarditis. Chest. 2007;132(3):1025-35.

11)   Yiu KH, Siu CW, LeeKL, et al. Emerging trends of community acquired infective endocarditis. Int J Cardiol.2007;121(1):119-22.

12)   Fedeli U, SchievanoE, Buonfrate D, et al. Increasing incidence and mortality of infective endocarditis:a population-based study through a record-linkage system. BMC Infect Dis. 2011;11:48.

13)   Sy RW, KritharidesL. Health care exposure and age in infective endocarditis: results of a contemporarypopulation-based profile of 1536 patients in Australia. Eur Heart J. 2010;31(15):1890-7.

14)   Lockhart PB, BrennanMT, Thornhill M, et al. Poor oral hygiene as a risk factor for infective endocarditis-relatedbacteremia. J Am Dent Assoc. 2009;140(10):1238-44.

15)   Chen SJ, Liu CJ, ChaoTF, et al. Dental scaling and risk reduction in infective endocarditis: a nationwidepopulation-based case-control study. Can J Cardiol. 2013;29(4):429-33.

16)   Ministry of Health.The New Zealand Oral Health Survey: Key Findings. Wellington: Ministry of Health;2010.

17)   Whyman RA, MahoneyEK, Stanley J. Admissions to New Zealand public hospitals for dental care: A 20-yearreview. Wellington: Ministry of Health; 2013.

18)   Gowda SS, ThomsonW, Foster Page LA, Croucher NA. Dental caries experience of children in Northland/TeTai Tokerau. N Z Dent J. 2009;105(4):116-20.

19)   Aung YM, Tin Tin S,Jelleyman T, Ameratunga S. Dental caries and previous hospitalisations among preschoolchildren: findings from a population-based study in New Zealand. N Z Med J. 2019;132(1493):44-53.

20)   Atkinson J, SalmondC, Crampton P. NZDep2018 Index of Deprivation, Interim Research Report. Wellington:University of Otago; 2019.

21)   Clark EG. Supervisedtooth brushing in Northland. Dunedin, New Zealand: University of Otago; 2017.

22)   Wauchop K, ShettyA, Bremner C. The epidemiology of acute rheumatic fever in Northland, 2012-2017.N Z Med J. 2019;132(1498):32-40.

23)   Stats NZ. 2018 CensusPlace Summaries - Northland Region. Wellington: New Zealand Government; 2018. Availablefrom: https://www.stats.govt.nz/tools/2018-census-place-summaries/northland-region.Accessed 17 May 2021.

24)   Steele JG, Walls AW.Strategies to improve the quality of oral healthcare for frail and dependent older people. Qual Health Care. 1997;6(3):165-9.

25)   Bai AD, Steinberg M, Showler A, et al. Diagnostic Accuracyof Transthoracic Echocardiography for Infective Endocarditis Findings Using TransesophagealEchocardiography as the Reference Standard: A Meta-Analysis. J Am Soc Echocardiogr.2017;30(7):639-46. e8.

26)   Durack DT, Lukes AS,Bright DK. New criteria for diagnosis of infective endocarditis: utilization ofspecific echocardiographic findings. Duke Endocarditis Service. Am J Med. 1994;96(3):200-9.

27)   Ministry of Health.Promoting Oral Health: A toolkit to assist the development, planning, implementationand evaluation of oral health promotion in New Zealand. Wellington: Ministry ofHealth; 2018. Available from: https://www.moh.govt.nz/notebook/nbbooks.nsf/0/f1f6c7e1b5c5c9f5cc257409006dbe70/$FILE/promoting-oralhealth-a-toolkit-jan08.pdf.

28)   Moore D, Poynton M,Broadbent JM, Thomson WM. The costs and benefits of water fluoridation in NZ. BMCOral Health. 2017;17(1):134.

29)   Hobbs M, Wade A, JonesP, et al. Area-level deprivation, childhood dental ambulatory sensitive hospitalizationsand community water fluoridation: evidence from New Zealand. Int J Epidemiol. 2020;49(3):908-16.

30)   NDHB. Position statement oncommunity water fluoridation. Northland: NDHB; 2016. Available from: https://www.northlanddhb.org.nz/assets/Your-Health/Northland-DHB-CWF-Position-Statement.pdf.Accessed 17 May 2021.

31)   de Silva AM, Hegde S, Akudo Nwagbara B, et al. Community-basedpopulation-level interventions for promoting child oral health. Cochrane Databaseof Systematic Reviews 2016,9. Art. No.: CD009837.

32)   Broughton JR, Maipi JTH, Person M, et al. Reducing diseaseburden and health inequalities arising from chronic disease among indigenous children:an early childhood caries intervention in Aotearoa/New Zealand. BMC Public Health.2013;13:1-8.

33)   Chen SJ, Liu CJ, Tze-Fan C, et al. Dental Scaling and RiskReduction in Infective Endocarditis: A Nationwide Population-Based Case-ControlStudy. Can J Cardiol. 2013;29(4):429-33.

34)   Infective EndocarditisProphylaxis Expert Group. Prevention of endocarditis. 2008 update from Therapeuticguidelines: Antibiotic version 13, and Therapeutic guidelines: Oral and dental version1. Melbourne: Therapeutic Guidelines Limited; 2008.

35)   Habib G, LancellottiP, Antunes MJ, et al. 2015 ESC Guidelines for the management of infective endocarditis:The Task Force for the Management of Infective Endocarditis of the European Societyof Cardiology (ESC). Endorsed by: European Association for Cardio-Thoracic Surgery(EACTS), the European Association of Nuclear Medicine (EANM). European Heart Journal.2015;36(44):3081.

Contact diana@nzma.org.nz
for the PDF of this article

Subscriber Content

The full contents of this pages only available to subscribers.
Login, subscribe or email nzmj@nzma.org.nz to purchase this article.

LOGINSUBSCRIBE