 
|
|
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Journal of the New Zealand Medical Association,
05-May-2006, Vol 119 No 1233
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Patients consulting outside of funded practices within
primary health organisations: implications for utilisation reporting
Jayden MacRae, Cathy O’Malley, Mary Brown
Wellington Independent Practice Association (WIPA) Limited
manages five primary health organisations (PHOs) in a defined geographical area
throughout the lower North Island of New Zealand. As part of a contractual
obligation to PHO reporting requirements, WIPA supplies utilisation data to the
Ministry of Health (MOH) and district health boards (DHBs). This reporting is
important as it is reasonable to assume that the MOH and DHBs may make policy
decisions that affect funding based on the data supplied.
Because of the way in which PHOs were initially set up,
there are three groups of patients dealt with in this paper. Two groups are
funded on a population basis, those being enrolled or registered at a particular
practice in a PHO—these are collectively referred to in this paper as
“funded” patients. The third group are those patients who are not
enrolled or registered at any practice in the PHO which they visit, and are
referred to as “non-funded” patients.
Initial analysis conducted at the start of this process
indicated that there were many consultations being made by patients outside of
the practice at which they were funded but still within the same PHO. In this
paper this is referred to as a “PHO Funded Encounter” and
collectively as “intra-PHO flow”.
This paper aims to investigate and describe the issues and
potential effects this may have on utilisation reporting. To understand how this
can occur, one must consider the definition of what is being reported and the
two alternative processes that may be undertaken to complete the reports. The
terms used for the two processes will be “matched” and
“unmatched” utilisation reporting.
Although some investigations have been made into why
patients change their primary carers in New Zealand1 and abroad,2,3 these
studies do not apply to the current PHO setting and address “real”
and “permanent” movement of patients. This study looks at the
artificial affect the definition and reporting methods have on volumes.
It is hoped that information presented may help the MOH and
DHBs consider the definition of what should be reported and empower other PHOs
making future decisions on utilisation reporting and analysis.
Reporting definitions and interpretationsThe contracts for the five PHOs managed by WIPA all define
utilisation reporting as being for “First Level Services delivered to
Enrolled Persons.”4–8 Although the specific term
“enrolled” is used, it has been interpreted as “funded”
(as a registered patient has only a temporary status over 3 years within the
PHO).
Furthermore there are two alternative ways of interpreting
the requirement: counting where only patient consultations occur at the practice
at which the patient is funded (“exclusive” definition) or where
patient consultations occur at any practice in the PHO at which the patient is
funded (“inclusive” definition).
PHOs are primarily focused on population-based health, and
funding is based on patients at this level.9–11 The inclusive definition
potentially gives a more complete perspective on healthcare being delivered in a
PHO to its funded population, thus making it a more desirable approach. However
it is problematic when practices seeing patients on a “casual” basis
are not aware that the patient is registered or enrolled somewhere else within
the same PHO.
This problem applies to PHOs that consist of more than one
practice. Of the five PHOs managed by WIPA, four consist of more than one
practice. This paper mainly analyses data from these four PHOs.
Matched utilisation reportingThe fundamental steps involved in matched utilisation
reporting are:
The defining step in this process is
the matching of data (c). It affects the way in which all other steps are
undertaken and its purpose is to determine the consults completed in any
practice for those patients that are funded in the PHO (either enrolled or
registered within the last 3 years).
To match the data and identify PHO-funded patients, it is
necessary to extract data at an encounter level, with patient identifiers (in
this case National Health Index [NHI]). It then needs to be loaded into a
database management system and matched against a PHO-funding database. Ideally
this is done on NHI but where this is not available for all patients,
second-level matching may be done with patient
date of birth and
family name. The purpose of the match
is to identify all patients funded in the PHO on the day of the encounters to be
included in the reporting.
Aggregation of data is then done on a basis of counting all
encounters for PHO-funded patients and assigning to the appropriate reporting
categories. This data is then formatted in the appropriate way ready to be sent
to the MOH or DHB.
The matched process allows the application of either the
inclusive or exclusive definition to utilisation reporting.
Unmatched utilisation reportingThe fundamental steps involved in unmatched utilisation
reporting are:
In this process, extraction of data
at the practice can take two basic forms: either encounter-level extracts or
aggregated extracts. Either form would include encounters for patients enrolled
or registered at the practice only, which implicitly means they are funded
within the PHO. The chosen method will dictate what is done to aggregate the
data at the PHO.
Where encounter-level data is provided, the PHO must
aggregate the data into the appropriate categories. If aggregated data is
supplied, the PHO only need sum the aggregate totals from each practice in each
reporting category. Alternatively if a combination is supplied, the encounter
data should be aggregated and summed with the already aggregated data.
There may be advantages in extracting aggregated data, as
the volume transported would potentially be less and processing is distributed
among practices, rather than centrally at the PHO. By not matching it, this
process potentially requires less resource at the PHO level.
The unmatched process allows reporting based only on the
exclusive definition of utilisation. Where an inclusive definition is applied,
this method would potentially under-report volumes.
MethodIn this analysis, 6 months of matched utilisation data
was used from five PHOs. It was extracted from the wider set of data collected
on a routine basis for utilisation reporting.
The routine collection method involves extracting data
on a monthly basis from practice patient-management systems, transporting it via
HealthLink to the PHO data warehouse, where it is loaded via an automated
software mechanism. The data warehouse is built as a Microsoft SQL Server 2000
database.
At present, not all practices supply utilisation data.
Those that do are all using the MedTech32 patient management system (PMS). Some
practices had incomplete data for several months, due to anomalies in automated
collection routines.
Two quarters of data was used, being the periods 1
April 2004 to 30 June 2004 and 1 July 2004 to 30 September 2004.
Because of the way in which data is stored in, and
consequently extracted from, the PMSs, duplicate entries for some encounters
exist in the data. This has been accounted for by de-duplicating data at the PHO
using the assumption that a patient can have only one encounter for a particular
day at a particular practice with a particular healthcare provider.
Any patient that has two or more encounters in a day at
the same practice by the same provider has the encounters counted only once. All
matching was done using NHI, so any patient without was not matched and
therefore counted as a casual encounter. Some
date of birth and all
encounter date fields included a time
component, which was truncated in all calculations.
PHO information was matched based on the practice in
which patients were funded. Data covered two complete quarter periods, and
patients could potentially be registered with different practices and PHOs in
each quarter. Because of this, the nature of the encounter (being for a funded
or casual patient) was determined based on the registration status of the
patient in the quarter of the encounter.
Data was manipulated in SQL Server and exported to
Microsoft Excel software for presentation. Each category required for
contractual reporting was used to group and analyse the information. Comparisons
were made of the number of encounters that occurred within practices where the
patient was recorded as funded and those being part of the PHO where the patient
was funded, but not the funded practice itself.
For the purpose of this paper these will be termed
funded practice and funded PHO encounters respectively. The latter are those
that may be potentially lost in unmatched reporting. All analysis of funded PHO
patients had Otaki PHO data excluded from analysis, as it consists of only one
practice and has no possibility of having intra-PHO patient flow.
Because of the relatively recent commencement of Care
Plus, this category was not analysed.
ResultsFactors affecting
analysis—The data collected from practices for each PHO showed a
coverage rate of 77.5–86.7% of the total available for the period. This is
considered sufficiently high for the analysis in this paper.
The completeness of NHI data was between 96.8–99.8%
for the total of 527,175 encounter records analysed.
Funded
encounters—Table 1 shows the number of encounters in each PHO
broken into non-funded, and funded groups. The Practice group are those
encounters where patients have been funded at the practice at which the
encounter occurred. The PHO group are those encounters where the patient is
funded in the PHO, but not at the practice in which the encounter occurred. The
% PHO column shows the percentage of funded encounters that are made up by the
PHO group, indicating those encounters that may potentially be lost in unmatched
reporting and this is summarised in Figure 1.
Gender—There
was little difference in the rates of PHO funded encounters in the gender groups
amongst PHOs, ranging from 0 to 0.4% differences between male and females.
Deprivation
quintile—Figure 2 shows the funded encounters by quintile. A
quintile of “0” indicates no quintile assigned for the
patient.
Table 1: Encounters (by funding)
 Figure 1. Percentage of funded encounters (by
PHO)
 Age
group—Figure 3 shows the funded encounters by age group. The age
groups are deliberately split into uneven year intervals in order to match those
used for utilisation reporting. The 15 to 24 years category has by far the
largest percentage of PHO encounters, being 11.5 %. The over 65 category is the
lowest on 2.3 %, with the 45 to 64 years category low on 3.2 %. The remainder of
the categories are between 6.1 and 6.9%.
Figure 2. Percentage PHO-funded and total-funded
encounters (by quintile)
 Figure 3. Percentage PHO-funded compared with
total-funded encounters (by age group)
 Ethnicity—Figure
4 shows funded encounters by ethnicity, in the groups used for utilisation
reporting. Of particular note is the relatively large number of ethnicities not
stated. In this case, this seems to have been caused by a large number of
miscoded or non-standard coding. This group has a high funded PHO encounter
percentage, followed closely by New Zealand (NZ) Maori and South East Asians
(7.8%, 6.4%, and 6.4% respectively).
Figure 4. Funded PHO-encounter percentages by ethnic
group
 Figure 5. Funded encounters by HUHC status
 High User Health Cards
(HUHCs)—Figure 5 shows that holders of a HUHC have only 2.2% PHO
encounters, compared with 5.1% for non-card holders.
Community Service Cards
(CSCs)—Figure 6 shows that CSC holders have a 5.4% funded PHO
encounter rate, marginally higher than non-holders.
Figure 6. % PHO-funded and total encounters (by CSC
status)
 N=No; Y=Yes
DiscussionFactors affecting
analysis—The range of the data analysed was from 77 to 87% complete
for PHOs. Data that was not analysed was not collected at the time due mainly to
errors in collection or practices not yet implemented with the automated
collection software. The majority of the latter group are non-MedTech32
practices. For all PHOs, the completeness of data is acceptably high for the
analysis made in this paper.
The decision to only match data based on NHI makes this
analysis susceptible to error where there are low rates of its reporting. The
range by PHO is from 96.8% to 99.8%. It is consistent with the registered rates
of 94.1% and 94.8% respectively reported by other studies.10,12
It is likely that the majority of those patients without NHI
fall into the “casual” encounter group, and given that the rates of
recording are so high, analysis should not be affected.
PHO-funded
encounters—The purpose of this paper is to analyse the impact of
intra-PHO flow on unmatched utilisation reporting. Using a unmatched process,
funded PHO encounters would not be reported. A matched process would detect all
encounters by funded patients.
Figure 1 shows the proportion of encounters that would be
not be reported for each PHO under a unmatched process, ranging from 1.7 to
7.6%.
Both Kapiti and Wairarapa have after-hours services as part
of their PHOs. It is possible this contributes to a higher level of funded PHO
encounters, as patients cannot access their funded practice out of hours. In
other PHOs (although there may be a similar rate of after-hours encounters) they
would be outside of the PHO.
Tumai PHO has the lowest rate (possibly due to services in
the Porirua region running to capacity, thus making it difficult to gain
appointments outside of one’s funded practice).
Category
analysis—The purpose of analysing discrete categories is to
determine if the intra-PHO flow affects them uniformly. If this is not the case
an unmatched process has the implication that particular pockets of population
may appear to be under-utilising PHO resources. It may also imply discontinuity
of service which possibly would imply that the
Primary Health Care Strategy12 is not
achieving its goals in some PHO populations. This has previously been theorised
by Kerse and Mainous.13
All categories with the exception of gender show differences
in rates, both across categories and also across PHOs. Of most interest and
potential concern are the high rates in ethnic and deprivation categories.
Deprivation quintile 5 has a very high rate of PHO
encounters, particularly in Kapiti but also in Wairarapa. Capital shows an
increased level, but not to the same degree as the other two PHOs which may be
again an indication of health-seeking behaviour in after-hours services.
Another possibility that may contribute to this pattern may
be due to debts being incurred by more deprived populations, seeking treatment
where they have little or no debt incurred at the time of consultation thereby
avoiding confrontation or payment requests. This could also account for the high
rate seen in CSC holders. Literature1–3 on the topic of why patients
change GPs does not support this view, but with the studies being questionnaire
based and the stigma associated with debt, respondents may be disinclined to
share this as a reason.
The NZ Maori, Pacific Island, and Asian ethnic groups
exhibit very high rates of PHO encounters especially in Kapiti and Wairarapa.
With the exception of NZ Maori in Wairarapa, these groups are in a minority in
comparison with other PHOs. After-hours service use can be one possible
explanation for this, where these ethnic groups may consult more often out of
hours. Alternatively it may be that the PHOs are not catering for the ethnic
needs of these populations and patients must actively seek specific services
within the PHO that do.
15–24 and 25–44 age groups show increased levels
of PHO encounters through all PHOs. In Kapiti, the <5 and 5–14 age
groups are also high. This again is in fitting with trends in ethnicity where
minority groups seem to have high rates. It may also be likely that these age
groups tend to consult after-hours clinics more often. In all PHOs, the
45–64 and 65+ groups have remarkably low rates, which may be a positive
indicator of continuity of care for the older population.
It is also positive to see that HUHC users have a
consistently low rate across all PHOs. This may indicate that they tend to move
around less and are getting a higher continuity of care, in line with
Primary Health Care Strategy
goals.
Access to after-hours services may be confounding other
trends in all categories, possibly due to work commitments preventing
normal-hours consultations, lack of transport in rural areas in ‘one-car
families’, or waiting for ailments to turn into more serious or urgent
issues. Several of these issues may affect younger working populations, which
could explain the high trends for the 15–24 and 25–44 age groups in
Kapiti and Wairarapa.
ConclusionSome PHOs and some populations in the data analysed would be
significantly under-reported if an unmatched process was followed. In several
cases these are populations that are being “targeted”, making the
issue more pertinent. Because this is different between PHOs, some may be
affected more than others, thus putting them on an uneven footing when
utilisation rates are compared at a regional or national level.
The implication of this is that if either an exclusive
definition of what is reported or an unmatched reporting process if followed,
utilisation rates may appear artificially low. If left unclear this may mislead
policy decisions at MOH and DHBs.
Although the inclusion of after-hours services within the
PHO is believed to be a major contributor to the trends seen in this paper, the
effect is not clear or proven. Therefore, further investigation may be important
in understanding the functioning of PHOs.
PHOs undertaking utilisation reporting should examine the
methods by which they are collecting and reporting information and consider the
potential impact some of the factors highlighted in this paper may affect them,
especially those with after-hours services or that have target age, ethnic, and
deprivation groups as minorities.
The MOH and DHBs should consider carefully the definition of
what should be included in utilisation reporting and the effect this may have on
data influencing policy and funding. Indeed, (in these early stages) PHOs would
be wise to consider having the inclusive definition applied where possible using
a matched reporting process to capture and illustrate a fuller picture of
utilisation.
Author information:
Jayden MacRae, Team Leader, Information Management; Cathy O’Malley, Chief
Executive Officer; Mary Brown, General Manager, Shared Services; Wellington
Independent Practice Association (WIPA) Limited, Wellington
Correspondence:
Jayden MacRae, WIPA, PO Box 27-380, Wellington. Fax: (04) 801 8715; email: jayden.macrae@wipa.org.nz
References:
|
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| | Current
issue | Search journal |
Archived issues | Classifieds
| Hotline (free ads) Subscribe | Contribute | Advertise | Contact Us | Copyright | Other Journals |
|