This thesis describes a study of the utilisation of computers by individual general practitioners (GPs) in Australia, and compares the practice behaviour of GPs who use a computer as a clinical tool, either by prescribing, ordering tests, or storing patient data in an electronic medical record format, with those who do not use a computer for these functions.
A survey of individual GP’s use of computers was conducted among 1,336 GPs who participated in the Bettering the Evaluation and Care of Health (BEACH) program between October 2003 and March 2005. The GPs were then assigned to groups according to their clinical use (or not) of a computer, and were compared on a range of variables including the characteristics of the GPs themselves, their practices, their patients, the morbidity they managed for their patients, and the managements they provided. Their behaviour was also compared, using a set of quality indicators designed for use with the BEACH data, and applicable in a primary care setting, to determine whether the clinical use of a computer has an affect on the quality of care GPs provide to their patients. Finally, GPs who use clinical software with embedded pharmaceutical advertising were compared with GPs not exposed to advertisements via this media, to determine whether such advertising influences the prescribing behaviour of GPs to favour advertised brands.
From 44 quality indicators examined, clinical computer users performed ‘better’ on four and ‘worse’ on four. For the remaining 36 they exhibited no difference. Exposure to pharmaceutical advertising embedded in clinical software did not influence the prescribing behaviour of the GPs so exposed.
Despite the belief espoused in the literature that computer use will improve the quality of patient care, I have found no evidence to demonstrate that the use of a computer for clinical activity has (as yet) affected, either positively or negatively, the quality of care GPs provide to their patients. The current push to computerise general practice will mean that this method of assessment will be difficult to replicate in the future, given the absence of control groups. Other research methods will need to be developed.