JMIR Publications

ABSTRACT

Background:

General practice electronic health records contain a wealth of patient information. However, these data were collected for clinical purposes. Hence, questions remain around the suitability of using these data for other purposes including epidemiological research, developing and validating clinical prediction models, conducting audits and informing policy. This study assessed the quality of data in Australian and Canadian general practice electronic health records in the context of osteoarthritis for the purpose of externally validating a clinical prediction model for total knee replacement surgery.

Objective:

To assess the quality of data in Australian and Canadian primary care electronic health records for the purpose of externally validating a clinical prediction model for use in patients with osteoarthritis.

Methods:

A data quality assessment was conducted on 201,462 patient general practice electronic health records from Australia provided by NPS MedicineWise, and 92,425 from Canada provided by the Canadian Primary Care Sentinel Surveillance Network. Completeness, plausibility and external validity of data elements relevant to osteoarthritis were assessed.

Results:

There were minimal incomplete and implausible data fields for age and gender (<1%), geographical location (<5%) and commonly co-occurring comorbidities (<10%) in both data sets. However, weight, height, body mass index and Canadian Index of Multiple Deprivation contained over 50% missing data. The recording of osteoarthritis by age and gender in both data sets were similar to national estimates, except for patients aged 80+ (Australia: 16.6% [95% CI 16.0%-17.3%] vs 13.1% [95% CI 11.2%-15.4%]; Canada: 36.7% [95% CI 36.1%-37.2%] vs 50.8% [95% CI 50.7%- 50.9%]. Total knee replacement rates were substantially lower in both electronic health record data sets compared with national estimates (Australia: 72 vs 218 per 100,000; Canada: 0.84 vs 200 per 100,000).

Conclusions:

Age, gender, geographical location, commonly co-occurring comorbidities and prescribing of osteoarthritis medications in Australian and Canadian general practice electronic health records are suitable for use in clinical prediction model validation studies. However, body mass index and Canadian Index of Multiple Deprivation are unfit for use due to large proportions of missing data. Rates of total knee replacement surgery were substantially underreported and should not be used for prediction model validation. Better harmonisation of patient data across primary and tertiary care is required to improve the suitability of these data. In the meantime, data linkage with national registries and other health data sets may overcome some of the data quality challenges in general practice EHRs.