Coordination of health care – experiences with GP care among patients aged 45 and over, 2016
This technical note summarises the methods used to calculate descriptive statistics and results against performance indicators presented in the report Healthy Communities: Coordination of health care – experiences with GP care among patients aged 45 and over, 2016.
Healthy Communities: Coordination of health care – experiences with GP care among patients aged 45 and over, 2016 presents information and statistics from the 2016 Australian Bureau of Statistics (ABS) Survey of Health Care (the Survey), at the national, state and territory, and Primary Health Network (PHN) level.
As well as reporting against a number of indicators of patient-reported experience of general practitioner (GP) care developed from themes in the Survey, the report informs the Performance and Accountability Framework (PAF) indicators below. These indicators sit within the PAF dimension of quality of care as measured by continuity and responsiveness:
- 184.108.40.206: Measures of patient experience
- 220.127.116.11: GP-type service use.
This technical note provides information about the data source and methodology used for the report Healthy Communities: Coordination of health care – experiences with GP care among patients aged 45 and over, 2016, as well as references for further information.
The Coordination of Health Care Study
The Coordination of Health Care Study (the Study) was developed to fill an important data gap and provide information on patients’ experiences of coordination of care across Australia. It examines coordination and continuity of care in detail, and provides nationally consistent and local-level information on experiences with health care providers.
The Coordination of Health Care Study is comprised of two components:
- the Survey of Health Care 2016 (the Survey) (national level results released 20 September 2017)
- integrating data for consenting participants with specific data items from the Medicare Benefits Schedule (MBS) and Pharmaceutical Benefits Scheme (PBS) data (including Repatriation Pharmaceutical Benefits Scheme (RPBS) data), together with hospitalisation data including visits to emergency departments and admissions to hospital.
This first PHN area-level publication is for the first component of the Study, and uses the Survey data to report on GP care experiences among patients aged 45 and over who saw a GP at least once in the year prior to selection of the sample.
The Survey is a postal survey that focuses on understanding the experiences of coordination and continuity of care among patients aged 45 and over who had at least one GP visit in the 12 months prior to the selection of the sample (November 2014 to November 2015). The Survey was designed to provide robust samples of people with high health care needs from each of the 31 PHN areas across Australia. The Survey oversampled those who had seen a GP 12 or more times in the preceding 12 months (defined as high users), in order to capture greater numbers of patients with complex and chronic conditions, and therefore experiences with multiple providers including hospitals, specialists, and allied health professionals.
The Survey collected data on patients’ experiences of coordination and continuity of care; information transfer; and access to care for a range of health care services, including:
- medical specialists
- imaging and pathology tests
- hospital admissions
- emergency department (ED) visits
- other health professionals (including physical health and emotional and psychological health professionals).
The Survey was in the field April–June 2016, conducted by the ABS. At that time, information was collected from individuals about their experiences with the health system in the 12 months prior to their completion of the Survey as well as demographic information.
For further information on the survey methodology refer to Explanatory Notes: Survey of Health Care, Australia, 2016 (ABS 2016)External link, opens in a new window.[http://www.abs.gov.au/AUSSTATS/abs@.nsf/Lookup/4343.0Explanatory%20Notes12016?OpenDocument].
Scope and coverage
The sample frame for the Survey was the Medicare Enrolment Database (MEDB). The sample was selected by the Department of Human Services in accordance with a stratification and allocation specified by the ABS.
The Survey included patients aged 45 and over who had at least one GP visit in the 12 months between November 2014 and November 2015. An estimated 3% of patients visited a GP in the 12 months prior to selection (23 November 2015) but did not see a GP in the 12 months prior to enumeration (April 2016 to June 2016).
The scope of the Survey excluded the following:
- patients who were not registered with Medicare
- patients who had only had GP transactions which were not billed through Medicare (for example through doctors who draw a salary and do not bill to Medicare)
- patients who were in active military service and obtained all their medical services through the military
- patients who may have seen a GP at least once in the 12 months prior to enumeration but did not visit a GP between November 2014 and November 2015.
Analysis was undertaken to estimate the proportion of the population that were in scope of the Survey using 2015–16 MBS data. An estimated 94.5% of people aged 45 and over in the 30 June 2016 estimated resident population had a GP attendance in the 2015–16 financial year.
The Survey was a self-administered paper questionnaire and data were collected by mail. In order to obtain maximum response, a four stage mail-out approach was used. For further information on the content of the mail outs refer to Explanatory Notes: Survey of Health Care, Australia, 2016 (ABS 2016)External link, opens in a new window.[http://www.abs.gov.au/AUSSTATS/abs@.nsf/Lookup/4343.0Explanatory%20Notes12016?OpenDocument].
There were 124,072 patients selected for this survey and 35,495 participated, giving an overall response rate of 28.6%. For the Survey, it is not possible to distinguish between non-response and sample loss. For example, a patient may have been selected to participate, but will not have received any survey materials due to an out-of-date address on the MEDB.
Respondents were assigned to geographic regions (for example, State, PHN) and Socio-Economic Indexes for Areas decile that corresponded to their reported home postcode.
Weighting of data
As the data were collected from a sample survey, they were weighted to infer results for the total in-scope population. The weight for each respondent was calculated based on their probability of selection within each strata and then adjusted to meet the known population totals of the inscope population on the MEDB. For more details on the strata and weighting methodology refer to Explanatory Notes: Survey of Health Care, Australia, 2016 (ABS 2016)External link, opens in a new window.[http://www.abs.gov.au/AUSSTATS/abs@.nsf/Lookup/4343.0Explanatory%20Notes12016?OpenDocument].
Comparability with other surveys
This is the first time this Survey has ever been conducted across Australia. While the Survey contains similar concepts to those found in other ABS surveys, such as the 2015–16 Patient Experience (PEx) and 2014–15 National Health Survey (NHS), comparisons should be undertaken with caution. The results of the Survey are broadly consistent when restricting NHS and PEx to the same scope. For further information on the comparison of the Survey results with PEx and NHS results refer to Explanatory Notes: Survey of Health Care, Australia, 2016 (ABS 2016)External link, opens in a new window.[http://www.abs.gov.au/AUSSTATS/abs@.nsf/Lookup/4343.0Explanatory%20Notes12016?OpenDocument].
Presentation of results
Geography and socio-demographic characteristics
This report presents information split by several geographies and socio-demographic characteristics. Information is reported at the national level, by state or territory and by PHN areas (Department of Health 2016).
Results are also reported by a number of socio-demographic characteristics, including:
- age group
- self-assessed health status
- remoteness structure, 2011 (ABS 2013a)
- socioeconomic group – Index of Relative Socio-Economic Disadvantage, 2011 (ABS 2013b)
- level of highest educational attainment
- main language spoken at home
- whether covered by private health insurance.
The results represent respondents’ perception of their health status and views on experiences of using the health care system. As these data are self-reported, respondents’ recall, perceptions and views should be considered when interpreting the results.
Results are expressed as the percentage of respondents who chose a specific response option or options for a question. The reported percentage is calculated as the number of people with a particular characteristic, divided by the number of people in the population of interest, multiplied by 100 (Table 1.1).
All percentages in this report are weighted estimates that use person weights allocated to each Survey participant by the ABS.
Table 1.1: Indicator description and calculation
|Type of measurement||Percentage, reported at one decimal place1|
|Calculation||(Numerator divide denominator) x 100|
|Numerator||The numerator was calculated as the sum of calibrated sample weights for patients with the characteristic of interest and who were enumerated within the particular PHN area.|
The denominator varies by Survey data item. The denominator for many Survey data items is all adults within the PHN area. For other Survey data items, the denominator may be a subpopulation of adults within the PHN area.
The denominator was calculated as the sum of calibrated sample weights for adults who were enumerated within the PHN area (see ‘Appendix A’).
|Protection of confidential data||Percentages are calculated based on counts that have been randomly adjusted by a small amount to avoid the release of confidential data.|
|Confidence intervals||As an indication of the accuracy of estimates, 95% confidence intervals were calculated using standard error estimates of the proportion.|
- All percentages at national level have a margin of error of less than 1, so percentages are reported to one decimal place.
Most data items include a ‘not stated’ category. This is to capture scenarios where the respondent was in the applicable population for that data item, but did not answer the question and there was not enough information to impute their answer to that question.
The inclusion of ‘not stated’ in the denominator poses an issue when reporting percentages. For example, the percentage of patients who selected ‘always/usually’ for whether usual GP or others in your usual place of care involve you in decisions about your health care and the complement of ‘sometimes/never’ will not add up to 100%. For this reason, and given the percentages of ‘not stated’ responses were below than 2%, it was decided to exclude ‘not stated’ in the denominator for these measures. Affected measures should be interpreted with caution. See ‘Appendix B’ for proportion of ‘not stated’ responses for the affected measures.
Reliability of estimates
Two types of error are possible in estimates based on a sample survey. These are non-sampling error and sampling error.
Non-sampling error may occur in any data collection and at any stage throughout the survey process. Examples include:
- non-response by selected persons
- questions being misunderstood
- responses being incorrectly recorded
- errors in coding or processing the survey data.
Sampling error occurs because a subset of the total population is used to produce estimates that are designed to represent the whole population. Sampling error can be reliably estimated, as it is calculated based on the methods used to design surveys.
As the estimates reported in Healthy Communities: Coordination of health care – experiences with GP care among patients aged 45 and over, 2016 are based on information obtained from a sample survey, they are subject to sampling error. That is, they may differ from proportions that would have been produced if all persons in Australia had been included in the Survey. Confidence intervals are presented, in addition to proportions, to indicate the range in which the population value (as compared with the statistic derived from respondent surveys) is likely to lie.
Confidence intervals are constructed using the estimate of the population value and its associated standard error. There is approximately a 95% chance (that is, 19 chances in 20) that the population value is within 1.96 standard errors of the estimated proportion. The 95% confidence interval is equal to the estimated percentage plus or minus 1.96 standard errors.
Suppression of estimates
Estimates were suppressed if there was the likelihood of a non-representative sample, that is, where the survey sample count in the PHN area was less than 100 persons. Suppressed data were annotated with ‘NP’—not published.
Estimates for people who did not answer the question for a particular population characteristic, for example, main language spoken at home or private health cover, and thus ‘not stated’ were also annotated with ‘NP’—not published. This was due to small numbers of people not answering these questions and the difficulty of interpreting these numbers. Totals are still reported to show the estimated number of people who did not respond to these questions.
Variation or difference in observed proportions might reflect only a random variation or difference. To assess whether differences between estimates are statistically significant—that is, that they are unlikely to be due to chance alone—a z-score for the difference in observed proportions was calculated using the following formula:z = (proportion 1) - (proportion 2) over the square root of (SE of proportion 1)^2 + (SE of proportion 2)^2
If z was greater than -1.96 or less than 1.96, the difference was not considered statistically significant at the 95% confidence level.
If z was equal to or less than -1.96 or equal to or greater than 1.96, the difference was considered statistically significant at the 95% confidence level.
The results presented are crude rates, which reflect the actual patient experiences in the community. However, note that there may be differences in the underlying distribution across age, socioeconomic groups or other factors that may explain differences in the crude rates between populations.
Age-standardisation is a method of removing the influence of age when comparing populations with different age structures. Due to potential differences in age structures between population groups being compared in this report, a sensitivity analysis was undertaken to examine the difference between crude and age-standardised proportions of indicators by PHN to identify if age-standardisation was required.
The method of age-standardisation considered was direct age-standardisation to the Australian estimated resident population, subset to people aged 45 and over, as at 30 June 2001.
Results showed that the differences between crude and age-standardised results were quite small. There were some small changes to rankings of PHNs when using age-standardised results compared with rankings based on crude results, but most PHNs stayed in the same decile within the rankings.
There are some small differences in the age structure of patients aged 45 years and over across PHNs, but the differences are much smaller than those for patients of all ages. The relationships between age and the patient-reported experience measures in this report are not as strong as for health outcome measures which led to small differences between crude and age-standardised rates.
Due to the small differences between crude and age-standardised rates and the focus of the report on patient-reported experience measures—for which age-standardisation is less appropriate—the estimates reported in the report are crude rates.
Figure 1.1 compares age profiles across PHNs in 2016.
Figure 1.1: Distribution of 10-year age groups compared across PHNs, patients aged 45 and over, 2016
|Code||Primary Health Network||Age group (years)|
|45 to 54||55 to 64||65 to 74||75 to 84||85+|
|PHN201||North Western Melbourne||37%||29%||19%||11%||4%|
|PHN801||Australian Capital Territory||36%||29%||21%||10%||4%|
|PHN105||South Western Sydney||35%||30%||20%||10%||4%|
|PHN101||Central & Eastern Sydney||35%||28%||20%||12%||6%|
|PHN203||South Eastern Melbourne||34%||28%||21%||11%||5%|
|PHN104||Nepean Blue Mountains||34%||31%||22%||10%||4%|
|PHN304||Darling Downs & West Moreton||32%||29%||23%||11%||4%|
|PHN306||Central Qld, Wide Bay & Sunshine Coast||30%||29%||24%||12%||5%|
|PHN108||Hunter New Eng. & Cent. Coast||29%||28%||24%||13%||6%|
|PHN106||South Eastern NSW||29%||29%||24%||13%||5%|