Colin Fischbacher Information Services Division (ISD) Risk of hospitalisation and death following prostate biopsy in Scotland: retrospective cohort study Colin Fischbacher Information Services Division (ISD)

Acknowledgements Co-authors: David Brewster, John Nolan, Siân Nowell, David Redpath and Ghulam Nabi. Funding: Supported by CSO Grant Number CZH/4/1039

Background In Scotland prostate cancer accounts for 21% of cancers diagnosed in men and is the second most common cause of death from cancer in men. A Cochrane review, based on a meta-analysis of five randomised trials, concluded that screening does not reduce prostate cancer-specific and overall mortality The harms associated with PSA screening include complications of diagnostic procedures Little information is available about the extent of this harm in “real world” practice

Data Linked acute hospital discharge (SMR01), outpatient (SMR00), cancer registration, and mortality records Men undergoing first ever prostate biopsy between 2009 and 2013 Scottish Index of Multiple Deprivation (SIMD) 2012 used as indicator of socio-economic position Charlson score based on primary diagnosis and prior inpatient bed days during 5 years before prostate biopsy (for bed days excluding the most recent 6 month period)

Data quality assessment We undertook linkage and comparison of hospital vs pathology records of prostate biopsies for one area of Scotland (Tayside) to check the completeness and accuracy of recording of prostate biopsy on hospital records We reviewed archived primary care records or pathology records of patients identified as dying within 30 days of prostate biopsy to confirm they all definitely had a prostate biopsy within 30 days of death

Methods We calculated crude rates of hospitalisation and death per 1000 patients within 30 and 120 days of prostate biopsy. We calculated indirectly standardised hospital admission ratios (SHR) and mortality ratios (SMR) with a general population standard, adjusting for age, sex, deprivation and calendar year. Follow-up was from date of prostate biopsy to 30/120 days after biopsy, or to date of death, whichever occurred first. We reviewed original death certificates for those who died within 30 days of a biopsy.

Data quality Data supplied by the Tayside pathology laboratory included 1681 records of prostate biopsy. Computerised hospitalisation data yielded only 508 records of prostate biopsy for patients treated in Tayside hospitals during the same period (2009-2013). Of these 508 records, 495 (97%) had a record of prostate biopsy within seven days of the corresponding pathology record, and 477 (94%) matched exactly for date of biopsy. Assuming pathology records to be the “gold standard”, and based on exact matching of dates, the sensitivity for detecting prostate biopsy using hospitalisation records was only 28% (477/1681).

Results: overall We identified 10,285 men who underwent first prostate biopsy between 2009 and 2013 492 were re-admitted and 14 died within 30 days of their biopsy 2,929 were re-admitted and 67 died within 120 days of their biopsy. Hospitalisation records included a slightly higher proportion of patients aged ≥80 years (11% vs 7%), but did not differ in terms of the distribution of deprivation categories. Across the whole of Scotland, it was verified that all 14 patients dying within 30 days of prostate biopsy (according to linked hospitalisation and mortality records), had indeed undergone prostate biopsy ≤30 days before death.

Results: readmissions Compared with the general population men undergoing biopsy were nearly three times (2.7; 2.4, 2.9) more likely to be admitted to hospital within 30 days of biopsy adjusting for age, deprivation and calendar year) They were four times (4.0; 3.8, 4.1) more likely to be re-admitted by 120 days.

25% readmitted within 120 days; more admissions among those with comorbidity

risk of readmission increases with increasing comorbidity, but raised risk of readmission even in those without prostate CA diagnosis

Results: deaths Compared with the general population men undergoing biopsy were 60% more likely to die within 30 days of their biopsy (SMR 1.6; 0.9 2.7) though this result is imprecise Men were nearly twice as likely to die (SMR 1.9; 1.5, 2.4) within 120 days of biopsy. Mortality was increased even when men with a subsequent prostate cancer diagnosis were excluded (SMR = 1.8; 1.2, 2.5) Mortality was not increased for men under 60 years of age (SMR 0.9; 0.3, 1.9).

non-sig increase in mortality at 30 days, significant at 120 days, excess risk rises sharply with age, present even in those without prostate CA

Potential for bias These data had universal population coverage with active data quality assurance for admissions and cancer. Comparison with laboratory data confirmed in 97% of cases that prostate biopsy had taken place We studied mainly hospitalised patients, who may be at higher risk than those undergoing outpatient biopsy However increased risk was present even in those without a subsequent prostate cancer diagnosis We had no information on complications of prostate biopsy that were managed exclusively in an outpatient or primary care setting

Conclusions In an unselected population of men undergoing first ever prostate biopsy, the risk of hospitalisation and death from any cause was higher than in the general population up to at least 120 days after the procedure. Although some hospitalisations and deaths seem likely to be associated with a subsequent diagnosis of prostate cancer, it is likely that some are associated more directly or indirectly with their prostate biopsy. 34% of these biopsies were carried out among men aged 70 years or older and 11% over 80 years. The higher rates of events in older patients and patients with co-morbidity emphasises the need for careful patient selection for prostate biopsy.