An examination of practice referral rates in relation to practice structure, patient demography and case mix.

: This paper examines the variation in practice referral rates from 53 practices over a twelve-month period. Data from the Second National Morbidity Study were used in relation to practice structure; age, sex and Social Class composition of the patient population; and the case mix by practice. The results show that the age of the patient is an important determinant of the probability of referral, whereas Social Class has little influence. Analysis of the practice-based data showed that practices were highly concordant in their referral activity across sex and age (greater or less than 45 years), sex and Social Class (manual or non-manual), and across chapters of the disease classification. This degree of concordance points to characteristics of the practices rather than patients, and their problems as the main source of variation in practice referral behaviour.