T Morton-Jones1, M Pringle. 1. Department of General Practice, Queen's Medical Centre, Nottingham.
Abstract
OBJECTIVE: To derive a predictive model for national prescribing behaviour in terms of basic morbidity and demographic factors. DESIGN: 24 demographic, morbidity, and practice factors were entered into a multiple regression analysis to determine the net ingredient cost per patient. SETTING: The 90 family health service authorities in England for 1989. RESULTS: For net ingredient cost per patient only two demographic factors (numbers of pensioners and the mobility of the registered population measured by list inflation) and two morbidity related factors (standardised mortality ratios and numbers of prepayment certificates issued) significantly contributed to a multiple regression model. This model explained 81% of the variation in net ingredient cost per registered patient between family health services authorities. The model also enabled a weighting factor of 4.6 (95% confidence interval 3.2 to 6.7) to be derived for the net ingredient cost for elderly patients (compared with the existing prescribing unit factor of 3). CONCLUSIONS: The model shows that variations in prescribing costs essentially reflect demand. It also suggests that the current prescribing unit value of 3 for patients aged 65 or more underestimates the extra costs of prescribing for elderly patients.
OBJECTIVE: To derive a predictive model for national prescribing behaviour in terms of basic morbidity and demographic factors. DESIGN: 24 demographic, morbidity, and practice factors were entered into a multiple regression analysis to determine the net ingredient cost per patient. SETTING: The 90 family health service authorities in England for 1989. RESULTS: For net ingredient cost per patient only two demographic factors (numbers of pensioners and the mobility of the registered population measured by list inflation) and two morbidity related factors (standardised mortality ratios and numbers of prepayment certificates issued) significantly contributed to a multiple regression model. This model explained 81% of the variation in net ingredient cost per registered patient between family health services authorities. The model also enabled a weighting factor of 4.6 (95% confidence interval 3.2 to 6.7) to be derived for the net ingredient cost for elderly patients (compared with the existing prescribing unit factor of 3). CONCLUSIONS: The model shows that variations in prescribing costs essentially reflect demand. It also suggests that the current prescribing unit value of 3 for patients aged 65 or more underestimates the extra costs of prescribing for elderly patients.