Use of multiple methods to determine factors affecting quality of care of patients with diabetes

Background. The process of care of patients with diabetes is complex; howev er, GPs are playing a greater role in its management. Despite the research evidence, the quality of care of patients with diabetes is variable. In ord er to improve care, information is required on the obstacles faced by pract ices in improving care. Qualitative and quantitative methods can be used fo r formation of hypotheses and the development of survey procedures. However , to date few examples exist in general practice research on the use of mul tiple methods using both quantitative and qualitative techniques for hypoth esis generation. Objectives. We aimed to determine information on all factors that may be as sociated with delivery of care to patients with diabetes. Methods. Factors for consideration on delivery of diabetes care were genera ted by multiple qualitative methods including brainstorming with health pro fessionals and patients, a focus group and interviews with key informants w hich included GPs and practice nurses. Audit data showing variations in car e of patients with diabetes were used to stimulate the brainstorming sessio n. A systematic literature search focusing on quality of care of patients w ith diabetes in primary care was also conducted. Results. Fifty-four potential factors were identified by multiple methods. Twenty (37.0%) were practice-related factors, 14 (25.9%) were patient-relat ed factors and 20 (37.0%) were organizational factors. A combination of bra instorming and the literature review identified 51 (94.4%) factors. Patient s did not identify factors in addition to those identified by other methods . Conclusion. The complexity of delivery of care to patients with diabetes is reflected in the large number of potential factors identified in this stud y. This study shows the feasibility of using multiple methods for hypothesi s generation. Each evaluation method provided unique data which could not o therwise be easily obtained. This study highlights a way of combining vario us traditional methods in an attempt to overcome the deficiencies and bias that may occur when using a single method. Similar methods can also be used to generate hypotheses for other exploratory research. An important respon sibility of health authorities and primary care groups will be to assess th e health needs of their local populations. Multiple methods could also be u sed to identify and commission services to meet these needs.