Family practitioners and other staff working in primary care require comprehensive and accurate data on patients at the point-of-care if they are to provide high quality health services to their patients. Electronic patient records are an effective method of achieving this objective, by dispensing with the need to use difficult to access, and often illegible, paper-based records. Hence, the implementation of electronic patient records in primary care is a key objective of many health care systems, including both the USA and UK.1 This reflects a growing recognition of the potential benefits of electronic records on the safety, quality and efficiency of healthcare. Electronic patient records underpin many information technology initiatives in primary care, such as screening for identifying patients at high risk of cardiovascular disease, call–recall systems for asthma and other long-term disease management programmes, computerized decision support systems for prescribing, electronic ordering of tests and electronic referral systems to secondary care. These are all, however, dependant on comprehensive and accurate coded data. There are known to be large variations in the accuracy and completeness of the clinical information stored in electronic patient records.2 In a systematic review, Thiru et al.3 identified 52 studies that examined data quality in electronic primary care records. Quality of data was measured in different ways, most commonly by comparisons of rates derived from the electronic records with an external standard. Prescriptions had the highest rate of recording, probably because prescribing is a core function of many electronic patient record systems. The recording of diseases (i.e. diagnoses) varied, with completeness generally highest for diseases with clear diagnostic criteria. Lifestyle and socio-economic data had lower rates of recording than prescription or diagnostic data. In another systematic review, Jordan et al.4 identified 24 studies that examined morbidity coding in primary care. Recording of consultations was generally high (typically greater than 90%), but assigning a morbidity code during each consultation was more variable (66–99% complete). Coronary heart disease was the most commonly assessed disease register in previous studies and completeness of recording was generally moderate (typically around 70%). Positive predictive value of coronary heart disease registers was generally high (typically around 83–100%). Other diseases that were examined (such as asthma and epilepsy) showed similar patterns of completeness of recording and positive predictive value of recorded diagnoses, but rates were generally lower than for coronary heart disease. Two recent papers in Family Practice also look at the issues of recording and coding of data in primary care. Pascoe et al.5 identified major omissions in the cancer diagnoses held by five general practices in Leeds, UK. The recording of diagnoses in primary care was less complete and, when a diagnosis of cancer was recorded, it was generally less detailed than in the data held by the Regional Cancer Registry. Soler et al.6 describe the progress of the International Classification of Primary Care (ICPC) in the 21 years since its introduction. The classification, now endorsed by the World Health Organization, has been translated into 22 languages. The wide use of the ICPC facilitates international comparisons of clinical practice and coding in primary care. For the time being, however, the use of Read codes as the UK's standard classification system in primary care makes comparisons with countries using ICPC difficult.7 One important conclusion of previous studies on the use of electronic patient records in primary care is that the completeness and accuracy of data entry relies mainly on the enthusiasm of family practitioners. There are currently no agreed reference standards for reporting data quality in primary care and this limits measurement of data quality in electronic patient records. Clinicians do understand the potential benefits from the use of electronic patient records in their practices, but also cite major barriers to their implementation.8 These include the capital cost of investment in information technology (this may be less of an issue in the UK where the capital costs are largely met by the NHS) and the workload implications. A second key area is the lack of standards that permit effective, accurate and timely exchange of electronic clinical data between healthcare providers. For example, a system that allowed data from regional cancer registries to be used to update electronic patient records in family practices could overcome many of the problems identified by Pascoe et al. about the coding of cancer in primary care. Hence, key areas for further work are the development and evaluation of data quality standards for use in electronic patient records; and the evaluation of methods for improving data quality. Another key area is understanding what aspects of electronic patient records appeal to family practitioners, contribute most to quality and safety of care, and why some aspects of the use of electronic records (for example, ‘Choose & Book’, the on-line outpatient appointment booking system in England's NHS) have been unpopular with clinicians. Other important areas that need to be addressed include minimizing duplicate data entry (for example, between hospitals, disease registries, clinical databases, and family practice records) by developing the means to accurately and efficiently facilitate sharing of clinical records. Another area that remains to be evaluated is the impact of allowing patients access to their electronic records and adding or correcting information in these records. This is starting to happen in the UK and elsewhere, but its impact on data quality, although likely to be beneficial, is not yet well understood. Concerns about data security need also to be taken seriously and safeguards need to be put into place to ensue that there is no unauthorised access or misuse of these increasingly comprehensive data. Electronic patient records offer enormous benefits, not only for patient care but also, when aggregated, for secondary analysis; and when linked with other health and social care datasets, for outcomes measurement, quality improvement, public health surveillance, and research. 9,10 These benefits cannot be fully realized without high quality data. Systems in which ‘free text’ natural language (reflecting clinicians thought processes) could be coded and used for additional functionality are still far in the future. Hence, the development of methods and incentives for significantly improving the coding of clinical data and data quality in electronic primary care records remain a priority for health care information technology programmes. DeclarationFunding: Connecting for Health Evaluation Programme (NHS CFHEP 001) Ethical approval: Not required Conflict of interests: None References2 , , . Use of Read codes in diabetes management in a South London primary care group: implications for the production of disease registers in primary care , , , vol. (pg. -)5 , , , , , . Identifying patients with a cancer diagnosis using general practice medical records and Cancer Registry data , , , vol. (pg. -)6 , , , . The coming of age of ICPC: celebrating the 21st birthday of the International Classification of Primary Care , , , vol. (pg. -)10 , , , , . Identifying undiagnosed diabetes: cross-sectional survey of 3.6 million patients’ electronic records , , , vol. (pg. -) |