Heather J. Litman, Research Scientist, New England Research Institutes, 9 Galen Street, Watertown, MA 02472, USA e-mail: email@example.com
To use the population-based data from the Boston Area Community Health (BACH) Survey to estimate the likely magnitude (prevalence) of urological symptoms in the USA population in 2025, as health-services researchers use projections of the likely magnitude of disease to inform decisions on the future allocation of health resources.
Age and gender-specific prevalence rates from BACH were combined with USA population projections to estimate the likely magnitude of lower urinary tract symptoms (LUTS) and symptoms suggestive of urine leakage, painful bladder syndrome (PBlS) and prostatitis (men only).
In total and accounting for overlapping symptoms, 52 million adults in the USA will have symptoms of LUTS, urine leakage, PBlS or prostatitis in 2025. These urological symptoms have a large impact on physical and mental aspects of quality of life, that is comparable to other chronic conditions.
The future magnitude of symptoms indicative of these four urological conditions might reach the current level of cardiovascular disease in the USA, which is considered ‘a modern epidemic.’ Our projections have important implications for medical education, training of healthcare providers, health-services research, and policy and patient education.
Health-services researchers often use projections of the likely magnitude of disease to inform policy decisions on the future allocation of health resources . The value of disease projections has been recognized for several decades . While projections are available for major conditions (e.g. cardiovascular disease and cancer), few such exercises have been undertaken in the field of urology (mainly because the population-based prevalence rates required for the projection of urological disease have not been available ).
Projections of the likely magnitude (prevalence) of urological symptoms are important because: (i) they help to guide the allocation of scarce health resources, as ideally there should be correspondence between the distribution of resources and the magnitude of urological symptoms in the population; (ii) they can motivate changes in the content of medical education, providing a strategic opportunity for advanced preparation of the health workforce to meet emerging health needs; (iii) they highlight the need for patient-education materials that are clinically accurate and socioculturally appropriate, providing reliable information on urological symptoms and informed help-seeking.
With newly available population-based data from the Boston Area Community Health (BACH) survey it is possible to estimate the likely magnitude (prevalence) of urological symptoms in the USA population in 2025. First, using BACH data, we present age- and gender-specific estimates of symptoms suggestive of four urological conditions, i.e. LUTS, urine leakage, painful bladder syndrome (PBlS), prostatitis (men only) for the USA; second, the rapid ageing of the USA population over the next two decades is described; third, age- and gender-specific prevalence rates from BACH are combined with USA population projections to estimate the likely magnitude of symptoms suggestive of LUTS, urine leakage, PBlS and prostatitis in the USA in 2025; fourth, the amount of bother associated with urological symptoms is discussed and their impact on quality of life (QoL) is contrasted with other major life-threatening chronic diseases. Implications of the emerging magnitude of symptoms of urological disease for health-services research, medical education and planning by healthcare providers are also discussed.
Litwin et al. discussed the importance of estimating the future magnitude of urological diseases and identified obstacles to doing so, including reliance on utilization data. Even more problematic is the dependence on already diagnosed cases; the BACH study instead recognizes that people suffer from the symptoms of urological disease and not from the disease labels which a provider might or might not attach to some combination of signs and symptoms. Therefore, to reliably estimate the magnitude of urological problems, we focus on self-reported symptoms rather than physician-diagnosed disease states.
The protocols and informed consent procedures were approved by the Institutional Review Board at New England Research Institutes; all participants provided written, informed consent. The BACH survey is a population-based, random-sample epidemiological study of a broad range of urological symptoms. The BACH multistage, stratified cluster sample of 5506 people (2301 men and 3205 women) was recruited from April 2002 to June 2005. Individuals from selected census blocks were chosen with approximately equal numbers of Black, White, and Hispanic men and women by age group (30–39, 40–49, 50–59, 60–79 years). Sampling took place in five batches, each a random subsample of the overall BACH survey . Of selected households, 36% were screened, 30% refused screening and 34% could not be contacted after at least 16 attempts. BACH eligibility criteria included: self-identified as Hispanic, Black or White, competent to sign informed consent and able to speak English or Spanish well enough to complete the survey. Interviews were completed with 63% of the screener-identified eligible individuals from the selected households. Because of design requirements, the BACH subjects had unequal probabilities of selection into the study. For analyses to be representative of the city of Boston, observations were weighted inversely proportional to their probability of selection into the study . Weights were further post-stratified to the population of Boston according to the 2000 Census.
We considered the magnitude of symptoms suggestive of four conditions: LUTS, urine leakage, PBlS and prostatitis (men only). LUTS are assessed using the AUA Symptom Index (AUA-SI) , a clinically validated multidimensional measure of urological symptoms, with excellent test-retest reliability ; it has been widely used in epidemiological studies of LUTS [8,9]. Although the AUA-SI was originally developed to quantify LUTS typically associated with BPH, it is not specific to BPH and has been applied to women , although not validated among women. Following common convention, LUTS was defined as an AUA-SI score of ≥ 8.
To define urine leakage, people who reported ‘yes’ to at least weekly or daily incontinence were counted as having symptoms of weekly urine leakage. Symptoms of urine leakage experienced at least weekly have high clinical relevance, are probably associated with bother and reduced QoL, and are most likely to result in help-seeking . Symptoms of weekly urine leakage have been proposed as a more reliable measure than less frequent leakage as they rely on recent information and therefore the questions are perhaps better understood by participants . Symptoms of PBlS are defined as symptoms of pain increasing as the bladder fills and/or pain relieved by urination (‘fairly often’ to ‘almost always’) of ≥ 3 months’ duration . Symptoms of prostatitis in men are defined as symptoms of any perineal or ejaculatory pain and a Chronic Pain Symptom Index pain score of ≥ 4 [14,15].
To illustrate the gender-specific change in the USA population between 2000 and 2025, the population size and percentage change by gender from the United Nations (UN) World population prospects is presented. The UN estimated and projected population distributions by gender and age groups for 2000 and 2025, respectively, are used to calculate the likely increase in the prevalence of urological symptoms in 2025 . The UN provides three population projections for the USA, based on different sets of assumptions about future trends in fertility, mortality and migration . We use the most conservative (lowest) projection for the USA population to present the best-case scenario for the future. To calculate the probable increase in the USA in the age-adjusted prevalence of urological symptoms between 2000 and 2025, we multiplied the population size in 5-year age groups (30–34, 35–39, . . . , 75–79 years) by the age group-specific prevalence rates, and then summed across the age groups to yield the prevalence values for each of the urological symptoms considered.
To assess ‘disease-specific’ QoL, bother was measured by the BPH Specific Interference with Activities Scale  which asks ‘during the last month, how often have urinary experiences, pain or discomfort in your pubic area interfered with __’ where ____ includes seven activities (e.g. driving for 2 h without stopping; going to the movies, shows, church, etc.) The scale, ranging from 0 to 28, has been validated in men, with good internal consistency and reproducibility ; it has not been validated in women, although the questions are not gender-specific.
Alternatively, to study QoL that was not specific to any disease (‘disease non-specific’), both the physical and mental aspects of QoL were assessed using the Medical Outcomes Study Short Form-12 (SF-12) health-related QoL instrument [18,19]. The two component score (CS) scales (physical, PCS12, and mental, MCS12) were validated, and have excellent test-retest reliability; the scales have been widely used and are useful in comparing the effects on QoL across disease states. In the general USA adult population (aged ≥ 18 years), both scales have a mean of ≈ 50 and a sd of 10 points, with higher scores indicating better QoL. A study of demographic and clinical variations in health status found the SF-12 not to vary by racial/ethnic group .
The impact of urological symptoms and other major chronic illnesses on health-related QoL by gender was compared by fitting separate linear regression models. For instance, to assess the impact of LUTS on the bother score, a model was fitted with LUTS as the independent variable and bother score as the dependent variable; similar models were fitted for each urological symptom. With the response variable of PCS12 in men, eight separate models with different independent variables (e.g. LUTS, urine leakage, PBlS, prostatitis, high blood pressure, diabetes, heart conditions, high cholesterol) were fitted. The process was repeated for PCS12 in women, MCS12 in men and MCS12 in women. Models were not adjusted by age or race/ethnicity, to allow for comparison of the regression coefficients. The regression coefficients denote the mean increase or decrease in QoL score for the symptom or illness considered compared to those without the symptom or illness (reference group). To account for the complicated study design, the models were weighted and conducted in SAS version 9.1 (SAS Institute, Cary, NC) and SUDAAN version 9.0.1 (Research Triangle Institute, Research Triangle Park, NC).
According to the most conservative population projection for the USA, there will be 32.1 million men and 36.8 million women aged 60–79 years in 2025, 32.1 million more than at present (15.3 million more men and 16.8 million more women; Table 1). It was estimated that in 2000, ≈ 22% of men and ≈ 25% of women were aged 60–79 years; the denominators used for these calculations were men aged 30–79 years (and women aged 30–79 years) so as to be the same as the 50-year age span covered in the BACH survey. These percentages (of those aged 30–79 years) are projected to increase substantially in 2025, when 33% of men and 36% of women are expected to be aged 60–79 years, increases of 91.4% and 83.5%, respectively.
Table 1. Gender-specific change by age group (30–39, 40–49, 50–59, 60–79 years) in the USA population between 2000 and 2025. From 
Age group, years
Population (millions) 2000
% change, 2025
Based on these UN data and the prevalence rates by gender and 5-year age group from the BACH survey, estimated prevalence of symptoms suggestive of LUTS, urine leakage, PBlS and prostatitis (men only) for 2000 and their projected prevalence in 2025 are presented in Table 2. Because the BACH survey obtained reports of the symptoms of different disease conditions (rather than already diagnosed cases), overlapping symptoms can be investigated. Thus, Table 2 presents the estimated prevalence of symptoms of at least one of the conditions (LUTS, urine leakage, PBlS or prostatitis) as adding the prevalence estimates for symptoms of each condition would lead to an overestimate of the number of people affected. In summary, over 52 million people are projected to have symptoms of at least one of the urological problems in 2025. In both men and women, the largest projected percentage increase from 2000 to 2025 is in symptoms of urine leakage (49% for men, 38% for women).
Table 2. The probable increase in the USA in the age-adjusted prevalence of urological symptoms between 2000 and 2025
Prevalence estimates used are from the BACH survey for ages 30–79 years. UN population estimates for 2000 and the most conservative (lowest) projections for 2025 are used. Results have been rounded to the nearest thousand.
At least one of the above
Figure 1a,b depicts the increase in the age-adjusted prevalence of urological symptoms for the USA between 2000 and 2025, comparing those with symptoms of LUTS only, urine leakage only, prostatitis only, PBlS only and a category for those having symptoms of two or more of these conditions. Among men in 2025 (Fig. 1a), it is projected that 52% (>13 million) with urological symptoms will have symptoms associated with LUTS only and 31% (>8 million) will have symptoms of two or more conditions, while 1% (<1 million) will have symptoms associated with PBlS only. Among women in 2025 (Fig. 1b), 48% (>12 million) will have urological symptoms associated with LUTS only and nearly a third (32%; >8 million) will have symptoms of two or more conditions, while 2% (>0.5 million) will have symptoms of PBlS only.
Recognizing that a high prevalence of symptoms does not automatically translate into a high level of burden for individuals, in addition to likely magnitude we also assessed the bothersomeness of urological symptoms and their impact on QoL compared with other major chronic diseases. Considering ‘disease-specific’ QoL, compared with people reporting no symptoms of LUTS, those reporting LUTS had over a 5-point mean increase in bother score (Fig. 2). The effect of symptoms of urine leakage on bother was similar to that of LUTS and symptoms of PBlS were slightly more bothersome; symptoms of prostatitis were the least bothersome of the four conditions. Among women, bother scores in general were ≈ 0.9 points higher than in men (P < 0.001); this effect remained for all urological symptoms considered.
Figure 3 summarizes results of the impact of the urological symptoms considered here on ‘disease non-specific’ QoL (physical and mental health scores). These results show that: (i) the symptoms suggestive of each condition, not surprisingly, have a large impact on QoL; (ii) the magnitude of their impact on QoL shows little difference by gender; and (iii) the impact of urological symptoms on both physical and mental health components of QoL is generally similar to the effects of several other major chronic conditions. Figure 3a illustrates that among men the impact on physical health due to symptoms of LUTS, urine leakage, PBlS and prostatitis is similar to that of high blood pressure and diabetes on physical health, and even higher than that of high cholesterol. Among women (Fig. 3b), the impact of symptoms of LUTS, urine leakage and PBlS on physical health is also roughly comparable to that of high blood pressure, diabetes, heart conditions and high cholesterol. Even more striking is the impact that symptoms of LUTS, urine leakage and PBlS have on mental health; the effects due to urological symptoms are much larger than the impact of other major chronic illnesses in both men and women (Fig. 3c,d).
The BACH sample was collected in the Boston area and incorporated similar numbers of the three major racial/ethnic groups in Boston, but was unable to include other racial/ethnic groups throughout the USA (e.g. Asian Americans). The BACH sample was compared to three different government-sponsored national surveys (the National Health and Nutrition Examination Survey, the National Health Interview Survey and the national Behavioural Risk Factor Surveillance System) on many sociodemographic and health-related variables. While there are a few differences, most of the BACH estimates are comparable, suggesting that the BACH rates (with appropriate adjustments) could be generally applicable to the USA as a whole.
Very many adults in the USA will be living with symptoms of urological diseases in 2025. Even with our conservative estimates, >42 million will have symptoms of LUTS, >18 million will have symptoms of urine leakage, >7 million men will have symptoms of prostatitis and >4 million are projected to have symptoms of PBlS. Overall, ≈ 52 million people are projected to have symptoms of LUTS, urine leakage, PBlS or prostatitis in 2025.
In addition to the impact that urological symptoms have on ‘disease-specific’ QoL, compared with other major often life-threatening chronic conditions, the urological symptoms considered here might have similar or even greater impact on ‘disease non-specific’ QoL. Urological symptoms represent a continuous, daily burden on social functioning (sleep, work, leisure activities and social relationships) whereas other major chronic diseases like diabetes, hypercholesterolaemia (which is often symptom-less and undiagnosed) and hypertension are characterized as ‘silent killers’; they might eventually have serious untoward consequences (death), but for prolonged periods are unrecognized and seldom a primary reason for healthcare use. In summary, symptoms suggestive of several urological conditions will probably increase to epidemic proportions in the USA in 2025, and while usually not life-threatening they constitute a burden in terms of their impact on an individual’s QoL. As such, they are likely to precipitate high use of healthcare resources and represent a future burden on the USA healthcare system.
The results we have reported have implications in several areas:
Medical education. The everyday care of patients with urological disease is still mainly provided by general internists and family practitioners. However, during medical education little attention is devoted to the diagnosis and management of urological disease. Epidemiological studies like BACH are revealing the complex multifactorial nature of symptoms of urological disease, which requires careful training in history-taking and sufficient time during the medical encounter to obtain a thorough patient history.
Healthcare providers. As most symptoms of urological disease are managed by primary-care providers, it is worrisome that this area of medical work is on the decline in the USA . If present trends persist there will be a shortage of primary-care providers in the USA by 2025, when the many people reported in this paper will require accurate diagnosis and long-term management. Given the lead-time required to correct trends, it is necessary that appropriate actions be initiated now (2007) so that urological symptom care can be adequately addressed as it emerges.
Health services research and policy. Some policy makers believe that utilitarian values (greatest good for the greatest number) should guide the allocation of healthcare resources in USA society. The projections of urological need presented here indicate that in 2025, about half of those Americans with urological symptoms will have LUTS only, about a third will have two or more symptoms and a relatively smaller proportion will have symptoms indicative of PBlS only. Knowledge of these disproportions can inform future decisions about the allocation of urological resources. Medical care will need to be organized, distributed and reimbursed in ways that promote quality, patient-centred urological care. Healthcare organizations will need to provide incentives and permit sufficient time for providers to conduct the type of examination and history-taking required for patients with urological symptoms.
Patient education. In 1996, US$791 million was spent on direct-to-consumer advertising in the USA, whereas currently an estimated US$4.1 billion a year is spent (Food and Drug Administration National Conference, November 2005). With the rapid ageing of the USA population, direct-to-consumer advertising often focuses on urological symptoms, advising the public to seek medical care and ‘Ask your doctor about . . .’ Indeed, ≈ 95 million Americans go on-line each year to obtain advice on their everyday health problems , yet some of the information is of questionable value or simply wrong . Clinically accurate educational materials need to be developed to alert patients about important urological symptoms and to provide guidance on appropriate help-seeking. Development of user-friendly materials will require sound qualitative research on sociocultural issues. The materials must also be available in different languages to overcome the cultural and psychosocial barriers associated with symptoms of urological disease in some groups.
Funding for the BACH Survey was provided by NIDDK (NIH) DK 56842. We also thank Carol Link and Raymond Rosen for their assistance.