 
|
|
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Journal of the New Zealand Medical Association,
04-May-2007, Vol 120 No 1253
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Doctors,
practices, patients, and their problems during usual hours: a description of
rural and non-rural primary care in New Zealand in 2001–2002
Phil Hider, Roy Lay-Yee, Peter Davis
The assessment of patient encounters with general
practitioners (GPs) at both rural and urban settings is a remarkably
under-researched area in New Zealand, despite the importance of this information
for the planning, funding, and delivery of health services and medical
education.
Most assessments of rural health services have focused on
providing a description of the characteristics of practitioners, their
workloads, or their distribution.1-3 These
studies have highlighted significant disparities in the availability and
distribution of GPs in New Zealand. Other examinations of the rural GP workforce
and their workload have used anonymous questionnaires sent to rural and
semi-rural GPs.4
In these surveys, rural GPs have identified that the lack of
locum relief, onerous on-call responsibilities, and rural GP shortages were all
important problems for rural health service
delivery.5
Without an equivalent to the National Minimum Dataset (NMDS)
that provides detailed attendance information about inpatient care,
community-based services have largely relied on occasional
audits6 and the findings from the RNZCGP
research group7 that has piloted an electronic
dataset amongst a number of practices to generate descriptive data about the
patients who receive care, their problems, and their treatments.
The National Primary Medical Care survey (NatMedCa) was
undertaken to describe primary health care in New Zealand, the first paper
provided an overview of the characteristics of providers and their
practices.8 This paper provides an analysis of
a representative sample of daytime attendances at rural and non-rural offices
around New Zealand and compares patients’ reasons for consultation,
morbidity managed by the GPs, the management of problems, referrals for
specialist care, and requests for investigations.
MethodsThe methods used to undertake the NatMedCa survey have
previously been described.8 In brief, the
survey was carried out during 2001/2 using a nationally representative,
multi-stage probability sample of GPs stratified by geographical location and
practice type. A sampling frame of active GPs was generated from telephone White
Pages listings.
Practices were defined as rural providers if their
rural ranking score was equal to or greater than 35. This is the criterion for
the Ministry of Health’s rural health benefits eligibility. The scale
allocates points on the basis of the frequency of on-call responsibilities;
requirement to be on-call for major trauma; the occurrence of regular peripheral
clinics; and the times required to travel to the nearest hospital, nearest
colleague, and the most distant boundary.
A follow-up survey of participating practices located
in country areas with a resident population less than 30,000, for which no score
had been provided, was carried out following data collection in order to
ascertain and confirm rural provider status according to the Ministry’s
criteria. In addition, to ascertain the population of rural provider
organisations a complete list was sourced via the Ministry.
GPs were asked to provide data on themselves, their
practice, and to report on their patients in each of two week-long periods
during usual working hours. The log questionnaire completed for all patients
seen during the data collection period, recorded demographic data. The visit
questionnaire recorded data on every 4th patient’s problem(s) and
management. Telephone calls and call-outs to visit patients or attend
emergencies were not included.
Statistical analyses were conducted taking into account
the stratified, multi-stage sampling scheme, the weights associated with each
stratum, and clustering at different sampling stages. Reason-for-visit and
diagnosis were coded using READ version 2 while drugs were classified using the
anatomically based Pharmacodes/ATC system.
Approval was obtained from ethics committees in all
areas represented in the survey. The study received advice from an advisory and
monitoring committee that included representation from consumers and relevant
professional groups.
ResultsData were contributed by 41 rural practices and 146
non-rural practices comprising 47 rural GPs and 197 urban doctors. Response
rates were similar between rural and non-rural providers (73.4% and 72.1%).
Table 1. Practice
characteristics
*Practice’s rural ranking score ≥35;
†.Percentages, means and 95% confidence intervals have been appropriately
adjusted to account for sample design. Missing data have been excluded from
calculations; ‡Combines both High User and Community Services Cards;
¶Significant difference between Rural and Non-rural practices.
Table 1 compares practice characteristics between rural and
non-rural locations. The mean number of full-time equivalent doctors and adult
patient fees were significantly lower at rural practices. Other differences
between rural and non-rural practices were not statistically significant. The
mean number of hours open per week was similar between rural and non-rural
practices, but rural venues provided more scheduled evening surgery hours and
fewer hours at weekends.
Rural practices offered more of almost all the services
listed in Table 1, including doctors providing maternity care, complementary
health, group health promotion, and community worker services. Computerised
patient records were especially common at rural practices (81.2% versus 67.4%).
In relation to governance, external management structure and
patient representation in management were uncommon at both settings,
particularly urban practices. The legal structure of most practices regardless
of location was as a sole trader business. Rural practices undertook more formal
needs assessment than their non-rural counterparts. Formal complaints policies
were relatively common at both rural and non-rural practices (62.3 and 59.2%
respectively), but few (31.6% and 30%) had written policy for quality
management. More rural practices were capitated, but fewer were budget-holding
practices.
Participating rural GPs were significantly more likely to
have graduated overseas than their non-urban colleagues. No other comparisons
between practitioners were statistically significant.
Rural and non-rural GPs were predominantly male, on average
they were 45 years old and had been in practice for an average of about 15
years. (Table 2). On average, rural GPs had been at their current practice for a
slightly shorter duration (8.8 versus 11.4 years). Fewer rural participants were
members of the RNZCGP (63.2% versus 81.8%) or the NZMA (44.8% versus 54.1%).
Rural participants worked on average more half-days per week
(8.2 versus 7.7) and they also saw more patients per week (on average 117 versus
99 patients).
The distribution of visits by people in the various age
groups was similar between rural and non-rural practices (Table 3). However,
rural providers had significantly more patients from deprived areas (NZDep01
1-3: 17.2% versus 32.4%), more Māori (20.2% versus 10.1%), but fewer
Pacific Islanders and fewer were judged to have language difficulties (only 1.2%
were not fluent in English).
The visiting profile of patients was the same (Table 4);
that is, the proportion new to the practice, or the doctor, and the average
number of previous visits over the last 12 months, were all similar. Most visits
at either rural or non-rural practices (86.0% and 89.3%) were financed by the
patient with or without general medical benefit subsidisation (GMS).
Visits funded by the Accident Compensation Corporation (ACC)
were significantly more common at rural practices (11.8% compared to 8.3% at
non-rural practices). The pattern of urgency for visits was also similar between
locales. Mean consultation length was marginally shorter at rural practices
(14.0 compared to 15.2 minutes).
Table 2. Practitioner
characteristics
*Practice’s rural ranking score ≥35;
†Percentages, means and 95% confidence intervals have been appropriately
adjusted to account for sample design. Missing data have been excluded from
calculations; ‡Significant difference between Rural and Non-rural
practices.
Table 3. Patient characteristics
*Practice’s rural ranking score ≥35;
†Percentages, means and 95% confidence intervals shown in all tables have
been appropriately adjusted to account for sample design. Missing data have been
excluded from calculations; ‡Ethnicity was self-reported, with multiple
categories allowed. One ethnic category was then assigned per patient according
to prioritisation of Māori and Pacific peoples. NZ Deprivation decile is a
measure of deprivation based on the area of residence; §Significant
difference between Rural and Non-rural practices.
Practitioners were asked to record up to four reasons for
each visit using, wherever possible, the patient’s own words. Problems
were determined by the practitioners. Non-rural patients had slightly more
reasons-for-visit or problems per 100 attendances than patients at rural
locations but neither difference was statistically significant.
Respiratory problems were the most common problem during
daytime at both rural and non-rural practices and accounted for 23.4 and 22.8%
of visits, at each location (Table 5). Injury was a significantly more frequent
problem among patients at rural practices compared with those in non-rural
areas.
Table 4. The visiting profile of
patients
*Categories are mutually exclusive, with maternity or ACC
taking precedence over cash/GMS where more than one is cited; †Up to 4
reasons could be recorded per visit; ‡Up to 4 problems could be recorded
per visit; §Significant difference between Rural and Non-rural
practices.
Overall,
rural GPs ordered significantly fewer tests and investigations compared to their
non-rural counterparts. Laboratory tests were less frequently requested at rural
practices while imaging tests, and other investigations (such as spirometry and
ECG tests) were ordered at comparable rates to urban practices.
Rural and non-rural practices were similar in relation to
the proportion of problems or visits that were associated with prescription or
other treatment items.
Follow-up and referral rates were also comparable at both
settings.
*Any number of prescription or other treatment items could
be recorded; †Up to 4 problems could be recorded per visit; ††
Follow-up and referral are not mutually exclusive; ‡ Significant
difference between Rural and Non-rural practices.
DiscussionThe National Primary Medical Care survey provides the most
comprehensive and representative sample of the character of rural and non-rural
office practice during usual working hours ever conducted in New Zealand. The
key impression is that there was a broadly similar pattern of practice and
practitioner characteristics as well as a presentation of patient problems
across locale during daytime in 2001/2.
Few significant differences between practices and GPs were
noted, although rural practices were staffed by significantly fewer GPs and
charged lower fees. Rural GPs were more likely to have graduated overseas.
Patient characteristics were also similar although patients at rural practices
were more likely to be Māori, reside in a lower decile area and be fluent
in English. The visiting profile of the patients were similar at each locale
although more visits were related to injury at rural areas and were funded by
the ACC.
Finally it also appears that rural practitioners generally
allocated similar treatment, follow-up and referral services as their urban
colleagues except they less frequently arranged laboratory tests and
investigations for patients.
A limitation of the survey is the absence of any data about
after-hours work or emergency call-outs. The greater burden of these activities,
especially for rural practitioners, have been associated with a considerable
amount of additional stress for the rural
workforce.9 The survey also does not provide
detailed information about the relative complexity of some presentations. For
example, the greater number of injuries managed in rural practice may sometimes
include the provision of X-ray and plaster services on-site.
READ codes provide some ability to accommodate symptoms that
may not have a recognisable pathological basis but are limited in their ability
to portray the complexity of the decision making processes needed to identify
different forms of psychological and physical ill-health in primary care.
Changes in primary care since 2002 may also limit the applicability of the data
to current practice. For example, since NatMedCa was undertaken many primary
care organisations have significantly increased the computerisation of their
practices and done much to facilitate patient
enrolment.10
There is a major problem in New Zealand associated with the
ongoing dearth of any reliable data about the morbidity encountered in primary
care. In contrast with New Zealand, extensive work has been undertaken in other
Western countries such as Australia11 and the
United States12 where databases detailing
patient encounters in primary care have been maintained for a number of years.
Comparisons between rural and non-rural general practice in
these countries have generally obtained results broadly consistent with those
documented by NatMedCa. They have described a similar pattern of slightly longer
working hours and more patient visits at rural compared to urban practices,
relatively higher rates of injuries treated at rural practices and a lower
number of visits by patients with mental
illnesses.13,14 Elsewhere these databases have
provided important information to assist workforce
planning15, promote service
development16, and support quality
improvement.17
The 2005 New Zealand Health Information
Strategy18 signals the electronic collection of
national primary care data as an ‘action zone’ priority for the next
five years. The results from NatMedCa provide a unique insight into the
information that could be available from a National Minimum Dataset in primary
care. Although the creation of such a dataset must first overcome a number of
problems related to resources and the compatibility of both software and
hardware the benefits of these data for health services planning, funding, and
delivery make this work a continuing
imperative.7
Future work should also aim to more fully describe the
workload of rural and non-rural practitioners particularly by capturing more
information about the complexity of patient care and the burden imposed by other
communication outside of the consultation including telephone calls and
administrative work as well as the requirements for being on-call and providing
after hours care.
Competing interests: None.
Author information: Phil Hider, Senior
Lecturer, Department of Public Health and General Practice, Christchurch School
of Medicine and Health Sciences, University of Otago, Christchurch; Roy Lay-Yee,
Research Fellow, Department of Sociology, University of Auckland, Auckland;
Peter Davis, Professor, Department of Sociology, University of Auckland,
Auckland
Acknowledgements: This study would not have
been possible without the generous assistance of all the participating general
practitioners, nurses, practice support staff, and their patients. The NatMedCa
study was funded by the Health Research Council of New Zealand. The assistance
of Alastair Scott, Antony Raymont, Peter Crampton, Sue Crengle, Daniel
Patrick, and Janet Pearson are also gratefully acknowledged.
Correspondence: Dr Phil Hider, Department
of Public Health and General Practice, Christchurch School of Medicine and
Health Sciences, University of Otago, P O Box 4345, Christchurch. Fax: ++64 3
364 3697; email: phil.hider@chmeds.ac.nz
References:
|
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| | Current
issue | Search journal |
Archived issues | Classifieds
| Hotline (free ads) Subscribe | Contribute | Advertise | Contact Us | Copyright | Other Journals |
|