Population-based linkage of health records to detect urological complications and hospitalisation following transrectal ultrasound-guided biopsies in men suspected of prostate cancer.

BACKGROUND

Transrectal ultrasound-guided prostate biopsy is a common urological procedure with known complications. We determined the urological complication and 30-day hospital admission rate in a population-based cohort using unique identifier and deterministic methodology of record linkage.

METHODS

715 men who underwent a transrectal ultrasound-guided biopsy in Tayside region of Scotland between April 2007 and September 2011 were identified from hospital records using their unique Community Health Index Number. Multiple hospital electronic databases (Central Vision, Insite, Wisdom, and TOPAS) and departmental-based resources (morbidity and mortality records, daily ward electronic records) were linked to estimate urological complications (urinary tract infection, urinary retention, haematuria) and rates of hospital admission. Cross-validation of information was obtained by searching a drug dispensing information network and microbiology database for infective complications. The hospital admission rate was compared for two different prophylactic antibiotic regimens used during the study period.

RESULTS

Of the 715 men who underwent transrectal ultrasound biopsy, 386 (53.9 %) were diagnosed with prostate cancer and 329 (46.1 %) were found to have benign histology. The hospital admission rate for urological complications within 30 days of the procedure for men without cancer was 1.95 % (14/715). The 30-day hospital admission rate was not different for different regimens of prophylactic antibiotics. There were 50 (6.99 %; 50/715) urine cultures requested to the microbiology department within 30 days of procedures; out of which 20 (2.79 %; 20/715) were positive. Most of these were generated from general practice requests. Eight blood cultures obtained within the same period were all negative. Eleven patients (1.7 %) presented with urinary retention during the same period and required indwelling catheterisation. None of them had any surgical procedure. Validation of electronic record linkage against telephonic questionnaires by specialist nurse showed a high reliability of the methodology (κ = 0.91).

CONCLUSION

High quality validated record linkage of cohort data in the present study using a unique identifier, protocol-based procedure and antibiotic prophylaxis showed that hospitalisation following TRUS biopsies occurs in less than 2 % of patients. However, a significant number of patients report to primary care, and centrally held records based on coding alone could underestimate the true incidence of complications.