Did the introduction of general practice fundholding change patterns of emergency admission to hospital?

To test the hypothesis that the introduction of general practice fundholding was associated with a change in the proportion of emergency admissions to hospital.

Before and after natural experiment with control group. The experimental group was first-wave fundholding general practices in the South Western Regional Health Authority, the control group was all practices that remained non-fundholding as of April 1993. Data were collected on episodes of care in hospitals in the South Western region involving cholecystectomy, hernia repair, intervertebral disc operation and prostatectomy. The additional impact of fundholding status on any underlying changes in proportions of emergency admissions was examined using multiple logistic regression.

There was no evidence of an interaction between fundholding status and before/after time period. Odds ratios and confidence intervals for the interaction of general practice fundholding status and time were: prostatectomy 1.02 (0.77 to 1.34); hernia repair 0.94 (0.7 to 1.24); intervertebral disk operations 1.67 (0.8 to 3.47); prostatectomy 0.94 (0.69 to 1.27).

The results provide no evidence that, in the first 2 years of the scheme, fundholding had an impact on the proportion of emergency admissions to hospital.