Pelvic organ prolapse (POP) is a common condition. Its incidence increases with age and its etiology is believed to arise from a combination of genetic and environmental risk factors.1 Although severe morbidity from prolapse is rare, in those women who seek medical care, surgical treatment may be offered if conservative treatment proves unsuccessful.2
Olsen et al3 in a seminal article, estimated that the lifetime risk of a prolapse or urinary incontinence (UI) operation in a U.S. health maintenance cohort was 11.1%. Recently Fialkow et al4 replicated this study using health maintenance organization data and confirmed Olsen's initial estimate, with a lifetime risk estimate of 11.8%. Olsen's estimate features heavily in the urogynecology literature and provides valuable background risk information to researchers and clinicians for the planning, delivery and evaluation of related gynecological services.
Despite the utility of these estimates, to our knowledge, no attempt has been made to estimate the lifetime risk of a prolapse operation in a general female population. Members of managed care organisations, which generally exclude older, socially disadvantaged and sicker members of the wider population,5,6 may not necessarily be representative for the purposes of drawing general inferences at a national level. Using the Western Australian Data Linkage System, which links together health databases on the entire, geographically defined population (2.1 million), we were able to identify all women who, between 1981 and 2005, underwent surgery for prolapse in any Western Australia hospital, irrespective of provider, facility or insurance status. A recent study found Western Australia to be the most representative of Australia's eight jurisdictions, making it well-placed to contribute health research that is applicable at a national level.7
Our objective was to perform a lifetime risk calculation for a developed, western country that was unaffected by the selection bias inherent in studies based on populations enrolled in particular insurance schemes. Our working hypothesis was that lifetime risk in our study would be higher than previous estimates.
MATERIALS AND METHODS
De-identified hospital morbidity and death data were extracted from the Western Australia Data Linkage System,8,9 a validated10–12 multi-set system that creates a linkage key between over 30 population-based administrative and research health data collections in Western Australia. We received hospital morbidity and death data for any female with an International Classification of Diseases (ICD) diagnosis or procedure code for pelvic organ prolapse on a hospital separation record. We defined our target case as a woman aged 18 years or older with a hospital separation where both a diagnosis and procedure for prolapse were recorded. Women with a prolapse diagnosis who underwent a hysterectomy in the absence of any other prolapse procedure were excluded from our case definition as we could not be certain from the linked data that the hysterectomy was performed primarily for prolapse. Owing to the structure of the linked data, we were able to distinguish true first-time procedures for prolapse from second and subsequent procedures. Only these first-time incident procedures were included in our analysis.
The year of surgery was classified into five calendar periods: 1981–1985, 1986–1990, 1991–1995, 1996–2000, and 2001–2005, and age at surgery was grouped into 5-year strata. Age-specific incidence rates were calculated using stratified procedure counts (by 5-year age group) and corresponding population denominators from the Australian Bureau of Statistics. We estimated lifetime risk based on the cross-sectional age-specific incidence rates in each calendar period It was derived by first calculating the cumulative incidence as CIi=1-e–IRi.ti in each of the age groups 18–19, 20–24, 25–29,…, 75–79, 80–84 years. Survival from prolapse to age 85 years was then given by:
whereof the final estimate of risk of prolapse through to age 85 years was derived from CI=1-S.
Age-standardized rates were calculated to assess the secular trend in surgery over two decades. Rates were standardized directly to the 2001 Western Australia population. Rate and risk calculations were performed using Microsoft Excel and STATA 10.0 for Windows. The project was approved for investigation by the human research ethics committee of the University of Western Australia.
During the period 1981–2005 there were 51,137 prolapse procedures performed in Western Australia hospitals and of these, the surgery was a first-time procedure in 44,728 women. Figure 1 shows the cumulative risk of surgery in the 1981–1985, 1991–1995 and 2001–2005 periods, based on the age-specific incidence rates in Table 1. Using these cross-sectional rates, the lifetime risk of undergoing a first-time procedure for prolapse by age 85 years was 20.5% (95% confidence interval [CI] 18.9–22.1%) in the 1981–1985 epoch, 21.1% (95% CI 19.8–125 22.6%) in the 1991–1995 epoch and 19.0% (95% CI 17.8–20.2%) for the period 2001–2005.
The age-standardized first-time procedure rate decreased 25% from 3.5 procedures in 1981 to a low of 2.6 procedures per 1,000 woman-years in 2005 (Fig. 2). The age-specific incidence rates increased with age, peaking at ages 45–49 years in 1981–1985 and at the later ages of 65–69 years in the periods from 1991–1995 to 2001–2005. The median age at surgery increased from 48.5 years in 1981–1985 to 53.3 years in the 5-year period to 2005.
The lifetime risk of POP surgery was 19% based on incidence rates in the 2001–2005 epoch. The rate of incident surgery decreased 25% over the 24-year study period, and the average age of patients undergoing surgery increased. The patterns of age-specific rates were comparable to previous reports3,4 of a peak in surgical intervention for POP in the 70–79 year age range, albeit that rates in Western Australia in the 1991 and 2001 epochs peaked earlier in the 65–69 year age group, a phenomenon consistent with what one expects in a population with higher intervention rates.
The incidence rate observed in Western Australia exceeded rates reported in several U.S. studies. Hamilton-Boyles et al13 report a rate of 1.5 procedures per 1,000 woman-years in 1997, whereas Babalola et al,14 using medical records from the Rochester epidemiology project, report an incidence rate of 1.3 per 1,000 woman-years in the period 1995 to 2002. Shah et al15 in a 2003 population-based study reported a higher rate of 1.8 per 1,000 woman-years. Considering that our case selection criteria were highly comparable to those in the Hamilton-Boyles study, our observed rate of 3.2 per 1,000 woman-years in 1997 far exceeded the rates in all of these reports, suggesting Western Australia had a high rate of prolapse surgery according to international comparisons.
From the late 1990s we observed a declining trend in the prolapse rate to a low of 2.6 per 1,000 woman-years in 2005, which was consistent with results from the U.S. studies.13,14 There are several possible reasons for the declining intervention rate. A change in the way prolapse is managed may have contributed to this trend. Despite the long-term availability of vaginal pessaries, it is possible that in recent years this nonsurgical treatment has become a more common first-line approach. This may be due to changes in the treatment seeking behaviors of patients' or a greater acceptability on the part of the specialist who, in the past, may have seen pessary treatment as an option only for poor surgical candidates.
It is also possible that over the study period the falling fertility rate in Western Australia16 and the increasing proportion of nonvaginal deliveries17 could have contributed to the decreasing prolapse rate. Although the etiology of prolapse remains poorly understood to a large extent, a small number of risk factors such as vaginal delivery and parity have gained much research support.18–22 The extent to which these factors could account for the falling prolapse rate via mechanisms linked to the protection of the pelvic floor is unknown and included here as a purely speculative comment deserving of further research. We hope that the next phase of our population-based study will examine these factors in detail.
What is more, as hysterectomy is considered a risk factor for a subsequent pelvic floor repair,23 it has been suggested that changing treatment patterns for hysterectomy may influence prolapse intervention rates. Using data from the Western Australia Data Linkage System, a recent study24 found that the hysterectomy rate in Western Australia decreased 23% over a 23-year period to 2003; therefore it is reasonable to suggest that a change in the prevalence of a history of hysterectomy may have influenced the downward trend in prolapse rates. Conversely, the rising prevalence of obesity25 and a recent upward trend in fertility levels26 are factors that could make a subsequent contribution to reversing this downward trend in the future.
Our results support the findings of Olsen et al3 and Fialkow et al4 that a relatively high percentage of the female population may expect to undergo surgery for pelvic floor disorders by their 80s. We found that the lifetime risk of surgery for pelvic organ prolapse by age 85 years was 19% based on rates in the 2001–2005 period. Our figure is considerably higher than the lifetime risk estimates reported previously.3,4 The difference could be explained by variations in the clinical criteria for prolapse intervention; differences in case definition or differences in the representativeness of the study population.
Our study was limited by the conventions applicable to administrative data that were not collected for the primary purpose of epidemiologic research. A lack of clinical detail limited finer specification of procedures than is presented here. It was, however, unlikely that our results were affected by coding artifacts or errors to a degree that could explain the apparently high rates. The requirements for both diagnostic and procedural criteria to satisfy the case definition limited the potential for over-enumeration. Previous work has shown that procedural information is one of the most reliable components of hospital separation data. Research from the Manitoba population health registry, to which the Western Australia data linkage system has a similar design, found that procedures were correctly identified in 98% of instances.27 Previous work using the Western Australia data linkage system has also found that administrative data provided a more complete ascertainment of surgical procedures for breast cancer than was possible through a clinical audit approach.11
To the extent that some misclassification may have occurred, it was likely to have led to an under-enumeration of cases where a procedure was recorded in the absence of an accompanying diagnosis code for prolapse. We did not exclude cases if they received concomitant treatment for other gynecological conditions such as leiomyoma or menstrual disorders, which would have resulted in additional cases in our study owing to this difference in case definition. To investigate the effect of this difference, we restricted our analyses to include only prolapse cases where the diagnosis for prolapse was the primary indication. We found that the risk estimate changed only marginally for the 1981–1985 and 1991–1995 5-year periods and decreased by 1% to 18% for the 2001–2005 period.
Although we included POP cases with concomitant UI procedures, we did not include cases of UI surgery in the absence of a POP procedure. This is an important distinction between our work and the previous studies3,4 that calculated a risk estimate based on surgery for POP, UI or both. Our results should be considered and interpreted in light of these differences.
Risk estimates based on cumulative incidence, which is conditional on survival to an advanced age, produce demonstrably higher figures than calculations based on multiple decrement life tables, which are not conditioned on survival to any particular age.28 However, this could not account for the high risk arising from our results, because the methods we employed were the same as those used in the previous studies.
Our relatively high result for the lifetime risk of surgery may simply represent the level of intervention typical of a population of females seeking medical care through a combination of public and private hospital facilities. Our state-wide population may be more representative of the typical level of need for surgical intervention than privately insured populations that tend to exclude both extremes of the social gradient; that is the most socially advantaged and disadvantaged. Whether women from these groups would be likely to have a greater prevalence of prolapse risk factors and thus a subsequently higher lifetime risk of surgery is again a further area of research that warrants investigation.
It appears that surgery for POP is common in Western Australia, evidenced by the high likelihood of surgery throughout the lifespan. Unquestionably, the burden of disease is even greater when considering the percentage of sufferers who never present for medical or surgical treatment. Understanding the complex etiology of this condition remains an important priority for urogynecology researchers, and prevention efforts need a greater evidence base if they are to be successful in reducing disease burden for what appears to be a high percentage of sufferers.
1.Altman D, Forsman M, Falconer C, Lichtenstein P. Genetic influence on stress urinary incontinence and pelvic organ prolapse. Eur Urol 2008;54:918–23.
2.Jelovsek JE, Maher C, Barber MD. Pelvic organ prolapse. Lancet 2007;369:1027–38.
3.Olsen AL, Smith VJ, Bergstrom JO, Colling JC, Clark AL. Epidemiology of Surgically Managed Pelvic Organ Prolapse and Urinary Incontinence. Obstet Gynecol 1997;89:501–06.
4.Fialkow M, Newton K, Lentz G. Lifetime risk of surgical management for pelvic organ prolapse or urinary incontinence. Int Urogynecol J 2008;19:437–40.
5.Fishman PA, H WE. Managed care data and public health: the experience of Group Health Cooperative of Puget Sound. Annu Rev Public Health 1998;19:477–91.
6.Schaefer E, Reschovsky JD. Are HMO enrollees healthier than others? Results from the Community Tracking Study. Health Aff 2002;21:249–58.
7.Clark A, Preen DB, Ng JQ, Semmens JB, Holman CD. Is Western Australia representative of other Australian States and Territories in terms of key socio-demographic and health economic indicators? Aust Health Rev 34:210–5.
8.Holman CDJ, Bass AJ, Rosman DL, Smith MB, Semmens JB, Glasson EJ, et al. A decade of data linkage in Western Australia: strategic design, applications and benefits of the WA data linkage system. Aust Health Rev 2008;32:766–77.
9.Holman CDJ, Bass AJ, Rouse IL, Hobbs MS. Population-based linkage of health records in Western Australia: development of a health services research linked database. Aust N Z J Public Health 1999;23:453–9.
10.Brameld KJ, Thomas MA, Holman CD, Bass AJ, Rouse IL. Validation of linked administrative data on end-stage renal failure: application of record linkage to a ‘clinical base population’. Aust N Z J Public Health 1999;23:464–7.
11.Watson N, Holman CDJ, Jamrozik K, Threlfall T, Kricker A. Validation of linked administrative data on primary clinical management of breast cancer in Western Australia in 1989. Perth: Centre for Health Service Research, Department of Public Health, University of Western Australia; 1997.
12.Teng TH, Finn J, Hung J, Geelhoed E, Hobbs M. A validation study: how effective is the Hospital Morbidity Data as a surveillance tool for heart failure in Western Australia? Aust N Z J Public Health 2008;32:405–7.
13.Hamilton Boyles S, Weber AM, Meyn L. Procedures for pelvic organ prolapse in the United States, 1979–1997. Am J Obstet Gynecol 2003;188:108–15.
14.Babalola EO, Bharucha AE, Melton LJ, 3rd, Schleck CD, Zinsmeister AR, Klingele CJ, et al. Utilization of surgical procedures for pelvic organ prolapse: a population-based study in Olmsted County, Minnesota, 1965–2002. Int Urogynecol J Pelvic Floor Dysfunct 2008;19:1243–50.
15.Shah A, Kohli N, Rajan S, Hoyte L. The age distribution, rates and types of surgery for pelvic organ prolapse in the USA. Int Urogynecol J. 2008;19:421–28.
16.Australian Bureau of Statistics. Births 2000. Canberra: ABS; 2001.
17.O'Leary CM, de Klerk N, Keogh J, Pennell C, de Groot J, York L, et al. Trends in mode of delivery during 1984–2003: can they be explained by pregnancy and delivery complications? BJOG 2007;114:855–64.
18.Swift SE. The distribution of pelvic organ support in a population of female subjects seen for routine gynecologic health care. Am J Obstet Gynecol 2000;183:277–85.
19.Hendrix SL, Clark A, Nygaard I, Aragaki A, Barnabei V, McTiernan A. Pelvic organ prolapse in the Women's Health Initiative: gravity and gravidity. Am J Obstet Gynecol 2002;186:1160–6.
20.Chiaffarino F, Chatenoud L, Dindelli M, Meschia M, Buonaguidi A, Amicarelli F, et al. Reproductive factors, family history, occupation and risk of urogenital prolapse. Eur J Obstet Gynecol Reprod Biol 1999;82:63–7.
21.Sze EH, Sherard GB 3rd, Dolezal JM. Pregnancy, labor, delivery, and pelvic organ prolapse. Obstet Gynecol 2002;100:981–6.
22.Samuelsson EC, Victor FT, Tibblin G, Svardsudd KF. Signs of genital prolapse in a Swedish population of women 20 to 59 years of age and possible related factors. Am J Obstet Gynecol 1999;180:299–305.
23.Mant J, Painter R, Vessey M. Epidemiology of genital prolapse: observations from the Oxford Family Planning Association Study. Br J Obstet Gynaecol 1997;104:579–85.
24.Spilsbury K, Semmens JB, Hammond I, Bolck A. Persistent high rates of hysterectomy in Western Australia: a population-based study of 83 000 procedures over 23 years. BJOG 2006;113:804–9.
25.Australian Institute of Health and Welfare. Health, wellbeing and body weight. Characteristics of overweight and obesity in Australia, 2001. Canberra: AIHW; 2004.
26.Australian Bureau of Statistics. Births 2008. Canberra: ABS; 2009.
27.Roos LL, Mustard CA, Nicol JP, McLerran DF, Malenka DJ, Young TK, et al. Registries and administrative data: organization and accuracy. Med Care 1993;31:201–12.
28.Schouten LJ, Straatman H, Kiemeney LA, Verbeek AL. Cancer incidence: life table risk versus cumulative risk. J Epidemiol Community Health 1994;48:596–600.