Abstract
Purpose:
More than 1 million prostate biopsies are performed annually among Medicare beneficiaries. Our objective was to determine the risk of serious complications requiring hospitalization. We hypothesized that with emerging multi-drug resistant organisms, there may be an increasing risk of infectious complications.
Materials and Methods:
In a 5% random sample of Medicare participants in SEER regions from 1991–2007, we compared 30-day hospitalization rates and ICD-9 primary diagnosis codes for admissions between 17,472 men who underwent prostate biopsy and a random sample of 134,977 controls. Multivariable logistic and Poisson regression was used to examine the risk and predictors of serious infectious and non-infectious complications over time.
Results:
The 30-day hospitalization rate was 6.9% within 30 days of prostate biopsy which was substantially higher than in the control population (2.9%). Adjusting for age, race, SEER region, year and comorbidities, prostate biopsy was associated with a 2.65-fold (95% CI,2.47–2.84) increased risk of hospitalization within 30 days, compared to the control population (p<0.0001). The risk of infectious complications requiring hospitalization after biopsy was significantly greater in more recent years (ptrend=0.001). Among men undergoing biopsy, later year, non-white race, and higher comorbidity scores were significantly associated with an increased risk of infectious complications.
Conclusions:
The risk of hospitalization within 30 days of prostate biopsy is significantly higher than in a control population. Infectious complications after prostate biopsy have increased in recent years, while the rate of serious non-infectious complications is relatively stable. Careful patient selection for prostate biopsy is essential to minimize the potential harms.
Keywords: prostate biopsy, complications, sepsis, infection, prostate cancer
INTRODUCTION
Prostate cancer is the most common non-cutaneous cancer in U.S. men1 and is primarily diagnosed through prostate needle biopsy. The age-adjusted rate of prostate biopsy per 100,000 Medicare beneficiaries for whites and blacks is 1580 and 1851, respectively.2 Furthermore, in the Medicare population, the risk of subsequent biopsy among men with a negative first biopsy was 12% at 1 year and 38% at 5 years.3
Prostate biopsies are performed transrectally with ultrasound guidance and when biopsies are taken an innoculum of bacteria from the rectum may be introduced into the prostate. Accordingly, infection is one potential risk of this procedure and can be severe requiring hospitalization.4 To reduce this risk, routine peri-procedural antimicrobial prophylaxis is utilized.
Numerous reports have demonstrated a rising prevalence of of antimicrobial-resistant organisms in the U.S. and abroad.5, 6 Notably, infectious complications that arise from antimicrobial-resistant organisms are more severe and costly to manage.7 We hypothesized that men undergoing prostate biopsy today may be at an increased risk of infectious complications than in the past. Other biopsy-related complications which may lead to hospitalization include hematuria, hematochezia, and urinary retention.8
Overall, the frequency of biopsy-related complications requiring hospitalization has not been described in the Medicare population. In light of recent controversies regarding prostate cancer screening, the incidence of infectious and other complications requiring hospitalization is useful to assess the costs and benefits of screening. Furthermore, given the frequency of this procedure, such results would also have significant public health implications.
METHODS
Medicare database
The study population was derived from a 5% random sample of male Medicare participants in Surveillance, Epidemiology and End Results (SEER) regions from 1991–2007. All individuals were aged ≥65 years, and enrolled in Medicare Part A and B for ≥1 month. We identified individuals who underwent ≥1 transrectal ultrasound-guided prostate biopsy based on carrier or outpatient claims. Prostate biopsy was identified using Current Procedural Terminology (CPT)/Healthcare Common Procedure Coding System (HCPCS) codes: 55700 (prostate needle biopsy), or both 76872 (transrectal ultrasound) and 76942 (ultrasound guided needle biopsy) together. For each individual, we only considered the first reported claim of a biopsy in the analysis. Thus, men with multiple biopsies are only present in the analysis once for their first reported biopsy. Men diagnosed with prostate cancer prior to this biopsy were excluded. As a control population, we identified a male sample aged ≥65 years in SEER regions enrolled in Medicare Part A and B for ≥1 month and selected a random day during their Medicare coverage as a time point for comparison.
Design
Similar to prior studies,9 we used ICD-9 codes for hospital admissions within 30-days of prostate biopsy or selection (controls) to examine primary diagnosis codes for biopsy-related infectious (kidney infection, urinary tract infection, prostatitis, cystitis, sepsis/bacteremia, endocarditis, hypotension, and postoperative infection) and non-infectious (gastrointestinal hemorrhage, hematuria, acute post-hemorrhagic anemia, and urinary symptoms/retention) complications (Appendix 1).
Statistical Analysis
Overall 30-day hospitalization rates were calculated by year from 1991–2007. We also calculated the rates of hospitalizations with infectious and serious non-infectious biopsy-related complications as primary diagnoses. T-tests and chi-square tests were used to compare baseline characteristics between biopsy and control groups. Multivariable logistic regression was used to examine the risk of hospitalization, infectious and non-infectious complications between the biopsy and control groups, adjusting for age, race, SEER region, year, and Charlson comorbidity scores.10, 11 In these models, region was classified as either midwest/central versus coastal based on prior studies demonstrating regional variation in antimicrobial resistance patterns.12 Multivariable Poisson regression models (including the same covariates) were used to examine predictors of serious infectious and non-infectious complication rates in the biopsy group. Subset analysis was also performed excluding men diagnosed with prostate cancer at or after the time of biopsy to avoid misclassification of early treatment-related hospitalizations or complications to the biopsy. Finally, 30-day mortality rates were examined in the biopsy and control populations. Multivariable analysis (using the same covariates as above) was performed to compare 30-day mortality rates between the biopsy and control groups, and between individuals with or without an infectious complication.
RESULTS
The median age was 73 years in the biopsy group and most were Caucasian (Table 1). Prostate cancer was diagnosed in 2,992 (17.1%) biopsied men. Biopsied men were slightly older, healthier and proportionally more were Caucasian than randomly-selected controls.
Table 1.
Demographics of the Men Undergoing Biopsy and Random Sample of Male Medicare Beneficiaries.
| Biopsy (n=17,472) | Control (n=134,977) | P-value | |
|---|---|---|---|
| Mean Age (years) | 73.0 | 72.4 | <0.0001 |
| Race (%) | <0.0001 | ||
| White | 14,308 (81.9%) | 106,980 (79.3%) | |
| Black | 1,444 (8.3%) | 10,231 (7.6%) | |
| Other | 1,720 (9.8%) | 17,766 (13.2%) | |
| Charlson comorbity score | <0.0001 | ||
| 0 | 16,226 (92.9%) | 122,561 (90.8%) | |
| 1 | 660 (3.8%) | 5,628 (4.2%) | |
| 2 | 340 (1.9%) | 3,266 (2.4%) | |
| 3+ | 246 (1.4%) | 3,522 (2.6%) | |
| Mean Year (SD) | 1998 (±5.0) | 2000 (±5.1) | <0.0001 |
| SEER Registry*, # (%) | <0.0001 | ||
| San Francisco | 1,273 (7.3%) | 14,639 (10.9%) | |
| Connecticut | 2,096 (12.0%) | 14,606 (10.8%) | |
| Detroit | 3,161 (18.1%) | 15,833 (11.7%) | |
| Hawaii | 488 (2.8%) | 4,947 (3.7%) | |
| Iowa | 1,842 (10.5%) | 13,627 (10.1%) | |
| New Mexico | 857 (4.9%) | 7,046 (5.2%) | |
| Seattle | 1,613 (9.2%) | 14,551 (10.8%) | |
| Utah | 1,004 (5.8%) | 6,168 (4.6%) | |
| Atlanta | 1,202 (6.9%) | 7,058 (5.2%) | |
| San Jose | 787 (4.5%) | 7,083 (5.2%) | |
| Los Angeles | 3,067 (17.5%) | 28,895 (21.4%) | |
| Rural Georgia | 82 (0.5%) | 524 (0.4%) |
To prevent bias in time-related analyses, we excluded SEER regions that began after 1992.
Overall, 1,209 (6.9%) men were hospitalized within 30-days after prostate biopsy, a significantly greater frequency than in controls (2.7%,p<0.0001). Infectious complications were the primary diagnosis in 67(0.38%) biopsy patients compared to 257(0.19%) controls (p<0.0001). Non-infectious biopsy-related complications were also significantly more frequent in the biopsy group (n=53,0.30%), than controls (n=53,0.04%, p<0.0001).
On multivariable analysis with age, race, SEER region, year, and Charlson comorbidity score, prostate biopsy was associated with a 2.65-fold increased risk of hospitalization within 30-days (95% CI 2.47–2.84,p<0.0001). Similar multivariable models showed that compared to controls, biopsy was associated with a significantly greater risk of hospitalization for infectious (OR 2.26, 95% CI,1.71–2.99,p<0.0001) and non-infectious complications (OR 8.48, 95% CI,5.68–12.64,p<0.0001). Subset analyses excluding men diagnosed with prostate cancer at or after biopsy yielded similar results: OR 2.32 (95% CI: 2.12–2.54, P<0.0001), OR 2.99 (95% CI: 2.20–4.04, P<0.0001), and OR 12.71(95% CI: 8.17–19.76, P<0.0001) for overall hospitalization, infectious and non-infectious complications, respectively, suggesting that prostate cancer treatment or treatment-related complications were not responsible for the increased risks of hospitalization and biopsy-related complications observed.
Figure 1 shows the frequency of overall hospitalizations, and hospitalization with infectious or non-infectious primary diagnoses. Throughout the study period, biopsied men had higher overall 30-day hospitalization rates than controls. The frequency of infectious complications increased over time among men undergoing biopsy (ptrend=0.001); whereas, the rate remained relatively stable in controls. Men undergoing biopsy also had higher rates of hospitalization for non-infectious complications compared to controls, with slightly more variability in the rates over time.
Figure 1.
30-day frequency of (a) overall hospitalizations, (b) hospitalizations with infection as the primary diagnosis, and (c) hospitalizations with non-infectious biopsy-related complications as the primary diagnosis in the biopsy (black circle) and control (gray squares) groups.
Among men undergoing biopsy, year, non-white race, and comorbidity score were significantly associated with the risk of infectious complications (Table 2). None of these covariates were significant predictors of non-infectious complications.
Table 2.
Factors associated with the risk of a primary infectious and non-infectious complication requiring hospitalization within 30 days of prostate biopsy.
| INFECTIOUS COMPLICATION OR (95% CI), p-value |
NON-INFECTIOUS COMPLICATION OR (95% CI), p-value |
|
|---|---|---|
| Age | 1.03 (0.99–1.07), p=0.11 |
1.04 (0.99–1.08), p=0.11 |
| Year | 1.09 (1.04–1.14), p=0.0006 |
1.05 (0.99–1.11), p=0.08 |
| Non-white race | 2.13 (1.28–3.57), p=0.004 |
0.92 (0.45–1.90), p=0.82 |
| Coastal region | 1.02 (0.61–1.69), p=0.94 |
0.83 (0.48–1.44), p=0.52 |
| Comorbidity score | 1.34 (1.06–1.71), p=0.017 |
0.92 (0.54–1.57), p=0.75 |
Finally, 55 (0.31%) men in the biopsy population died within 30 days, compared to 1474 (1.09%) controls. After adjusting for age, race, SEER region, year and comorbidity index, biopsied men had a reduced risk of mortality within 30 days compared to the control population (OR 0.29, 95% CI 0.22–0.38, p<0.0001), which was fairly constant over time. However, persons diagnosed hospitalized with an infectious complication had an increased risk of death within 30 days compared to those without hospitalization for an infectious primary diagnosis overall (OR 12.02, 95% CI 8.59–16.80, p<0.0001).
DISCUSSION
Prostate cancer screening is widespread in the U.S. and abroad. Following an abnormal screening, prostate biopsy is used to obtain a histologic diagnosis, and more than 1 million prostate biopsies are performed annually in the Medicare population.3 Although this procedure is typically considered benign, our results indicate that in the Medicare population, the rate of hospitalization within 30-days of prostate biopsy was more than double that of a control population. If this difference was similar to what would be found in a randomized trial, the number needed to harm would be 24 (i.e. 1 additional hospitalization within 30 days for every 24 prostate biopsies). Admissions primarily resulted from either exacerbations of chronic conditions or new issues, which included both infectious and non-infectious complications.
The issue of prostate biopsy complications is particularly germane in the context of the ongoing controversy regarding PSA screening.13, 14 Additionally, prostate biopsy is also used in other settings, such as monitoring for prostate cancer progression during active surveillance15. Thus, it is critical to establish morbidity rates associated with this procedure and for whom the risk of biopsy-related complications is greatest.
We observed an increase in biopsy-related infectious complications over the study period. A likely explanation for the rise in infectious complications is increasing antimicrobial resistance. The American Urological Association recommends antimicrobial prophylaxis for all patients undergoing prostate biopsy, with fluoroquinolones considered the antimicrobial of choice based upon randomized studies from the 1990’s.16–18
Nevertheless, fluoroquinolone resistance has increased during the past decade. In the US, fluroquinolone resistance in E. Coli bloodstream isolates increased from 0 to 12% from 1998–2007.5 Temporal increases in fluoroquinolone resistance have also been reported internationally.6, 19 These trends have been possibly linked to the introduction of these agents into animals raised for human consumption.20 We were unable to examine prophylactic antibiotic use in this study. Regardless, the increasing number of serious infectious complications over the study interval highlights the need to reevaluate the optimal prophylactic regimen for prostate biopsy and other invasive procedures. In the future, an assessment of antibiotic susceptibility with rectal swab cultures might be a cost effective strategy for targeted prophylaxis.
Although serious non-infectious urologic complications were substantially higher after biopsy than in the control group, the frequency of these complications remained relatively stable over time. This finding is not surprising, given that the technique of performing prostate biopsy transrectally using ultrasound guidance has remained stable during the study period.
Despite the increased frequency of complications within 30 days, the biopsy population actually had a lower 30-day mortality rate than controls after adjustment for demographic variables. This likely reflects improved patient selection for healthier men for prostate biopsy in that early mortality was reduced in this group. Nevertheless, individuals with an infectious complication were significantly more likely to die within 30 days, demonstrating the potential serious sequelae of a major infectious complication.
A limitation of our study is that individual-level covariate data regarding indications for biopsy, physician volume, number of biopsy cores, culture data, type and compliance with antimicrobial prophylaxis were unavailable, which would have been useful to help identify possible preventative strategies. Further study is warranted to determine whether increased bacterial burden (ex: increasing number of biopsy cores from 6 to 12) or inappropriate prophylaxis contributed to the rising rate of complications.
Another limitation is the possibility of misattribution of complications unrelated to prostate biopsy. To control for this, we studied the “baseline” 30-day hospitalization rate for a random sample of Medicare participants, and found it to be substantially lower. Although we excluded men with prevalent prostate cancer prior to biopsy, it is possible that some hospitalizations/complications from early treatment of prostate cancers diagnosed at this biopsy could have been misattributed to the biopsy itself. However, subset analysis excluding all men diagnosed with prostate cancer yielded similar results, suggesting that admissions for radical prostatectomy were not responsible for the trends observed. In addition, this would be unlikely to explain the dramatic increase in infectious complications over time.
Finally, although conditions resulting in hospitalization are generally more serious than those managed on an outpatient basis, admission criteria may vary between institutions. By focusing exclusively on primary diagnosis codes for inpatient hospitalizations, we may have underestimated the frequency of biopsy-related complications, and did not capture complications leading to return office or emergency room visits.
Despite these limitations, to our knowledge this is the first study to examine contemporary hospitalization rates after a “simple” outpatient procedure such as prostate biopsy in the Medicare population. Indeed, our study revealed a higher rate of serious complications following prostate biopsy than previously suggested. Although these findings may not be generalizable to younger men, it could be used to counsel men ≥65 years undergoing prostate biopsy on their risk of complications. Further studies are necessary to examine complications after prostate biopsy in younger men.
In the context of current health care reform, the economic implications of potential increased costs associated with biopsy-related complications also require further consideration. If similar results were found in a randomized trial, we estimated one additional hospitalization for every 24 biopsies performed. These results suggest the importance of better patient selection for prostate biopsy, and the identification of individuals at highest risk for complications.
Overall, the increasing frequency of serious infection associated with prostate biopsy, also recently demonstrated in a Canadian population,21 may suggest the need to re-examine protocols for peri-procedural antibiotic prophylaxis. Moreover, although prostate biopsy is often considered a “benign” procedure, these findings highlight the importance of individualized assessment of the risk-to-benefit ratio of this potentially risky procedure.
In conclusion, there is a non-trivial risk of serious complications after prostate biopsy. Overall 6.9% of men were hospitalized within 30-days of biopsy, compared to a baseline 30-day rate of 2.7% in randomly selected controls. After demographic adjustments, prostate biopsy was associated with a significantly increased risk of hospitalization, serious infectious and non-infectious complications. Finally, the risk of post-biopsy infectious complications increased over time. Together this highlights the importance of careful patient selection and counseling regarding prostate biopsy.
Acknowledgements
This study was supported in part by the Intramural Research Program of the Division of Cancer Epidemiology and Genetics. This study used the linked SEER-Medicare database. The interpretation and reporting of these data are the sole responsibility of the authors. The authors acknowledge the efforts of the Applied Research Program, NCI; the Office of Research, Development and Information, CMS; Information Management Services (IMS), Inc.; and the Surveillance, Epidemiology, and End Results (SEER) Program tumor registries in the creation of the SEER-Medicare database.
EMS is supported by Howard Hughes Medical Institute, the National Institutes of Health, AUA Astellas Rising Star Award and the Patrick C Walsh Prostate Cancer Research Foundation.
Footnotes
Previous Presentation: American Urological Association Meeting 2010
REFERENCES
- 1.American Cancer Society. Cancer Facts & Figures 2009 http://www.cancer.org/downloads/STT/500809web.pdf. Accessed October 7, 2009.
- 2.Prostate Cancer. In: Litwin MS and Saigal CS, editors. Urologic Disease in America US Dept of Health and Human Services, Public Health Service, NIG, NIDDK. Washington DC: US Government Publishing Office, 2007; NIH Publication No. 07–5512 [p.76]. [Google Scholar]
- 3.Welch HG, Fisher ES, Gottlieb DJ et al. : Detection of prostate cancer via biopsy in the Medicare-SEER population during the PSA era. J Natl Cancer Inst, 99: 1395, 2007 [DOI] [PubMed] [Google Scholar]
- 4.Ozden E, Bostanci Y, Yakupoglu KY et al. : Incidence of acute prostatitis caused by extended-spectrum beta-lactamase-producing Escherichia coli after transrectal prostate biopsy. Urology, 74: 119, 2009 [DOI] [PubMed] [Google Scholar]
- 5.Al-Hasan MN, Lahr BD, Eckel-Passow JE et al. : Antimicrobial resistance trends of Escherichia coli bloodstream isolates: a population-based study, 1998–2007. J Antimicrob Chemother, 64: 169, 2009 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Laupland KB, Gregson DB, Church DL et al. : Incidence, risk factors and outcomes of Escherichia coli bloodstream infections in a large Canadian region. Clin Microbiol Infect, 14: 1041, 2008 [DOI] [PubMed] [Google Scholar]
- 7.Roberts RR, Scott RD 2nd, Hota B. et al. : Costs attributable to healthcare-acquired infection in hospitalized adults and a comparison of economic methods. Med Care, 48: 1026. [DOI] [PubMed] [Google Scholar]
- 8.Rodriguez LV, Terris MK: Risks and complications of transrectal ultrasound guided prostate needle biopsy: a prospective study and review of the literature. J Urol, 160: 2115, 1998 [DOI] [PubMed] [Google Scholar]
- 9.Lawthers AG, McCarthy EP, Davis RB et al. : Identification of in-hospital complications from claims data. Is it valid? Med Care, 38: 785, 2000 [DOI] [PubMed] [Google Scholar]
- 10.Charlson ME, Pompei P, Ales KL et al. : A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis, 40: 373, 1987 [DOI] [PubMed] [Google Scholar]
- 11.Romano PS, Roos LL, Jollis JG: Adapting a clinical comorbidity index for use with ICD-9-CM administrative data: differing perspectives. J Clin Epidemiol, 46: 1075, 1993 [DOI] [PubMed] [Google Scholar]
- 12.Zhanel GG, Hisanaga TL, Laing NM et al. : Antibiotic resistance in outpatient urinary isolates: final results from the North American Urinary Tract Infection Collaborative Alliance (NAUTICA). Int J Antimicrob Agents, 26: 380, 2005 [DOI] [PubMed] [Google Scholar]
- 13.Schroder FH, Hugosson J, Roobol MJ et al. : Screening and prostate-cancer mortality in a randomized European study. N Engl J Med, 360: 1320, 2009 [DOI] [PubMed] [Google Scholar]
- 14.Andriole GL, Grubb RL 3rd, Buys SS et al. : Mortality results from a randomized prostate-cancer screening trial. N Engl J Med, 360: 1310, 2009 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Carter HB, Kettermann A, Warlick C et al. : Expectant management of prostate cancer with curative intent: an update of the Johns Hopkins experience. J Urol, 178: 2359, 2007 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Wolf JS Jr., Bennett CJ, Dmochowski RR et al. : Best practice policy statement on urologic surgery antimicrobial prophylaxis. J Urol, 179: 1379, 2008 [DOI] [PubMed] [Google Scholar]
- 17.Kapoor DA, Klimberg IW, Malek GH et al. : Single-dose oral ciprofloxacin versus placebo for prophylaxis during transrectal prostate biopsy. Urology, 52: 552, 1998 [DOI] [PubMed] [Google Scholar]
- 18.Aron M, Rajeev TP, Gupta NP: Antibiotic prophylaxis for transrectal needle biopsy of the prostate: a randomized controlled study. BJU Int, 85: 682, 2000 [DOI] [PubMed] [Google Scholar]
- 19.Peralta G, Sanchez MB, Garrido JC et al. : Impact of antibiotic resistance and of adequate empirical antibiotic treatment in the prognosis of patients with Escherichia coli bacteraemia. J Antimicrob Chemother, 60: 855, 2007 [DOI] [PubMed] [Google Scholar]
- 20.Heuer OE, Kruse H, Grave K et al. : Human health consequences of use of antimicrobial agents in aquaculture. Clin Infect Dis, 49: 1248, 2009 [DOI] [PubMed] [Google Scholar]
- 21.Nam RK, Saskin R, Lee Y et al. : Increasing hospital admission rates for urological complications after transrectal ultrasound guided prostate biopsy. J Urol, 183: 963. [DOI] [PubMed] [Google Scholar]