|Home | About | Journals | Submit | Contact Us | Français|
To examine the prevalence, natural history, and clinical significance of elevations in postvoid residual (PVR) volume in ambulatory older women
Prospective cohort study
Group health plan in Washington state
987 ambulatory women aged 55 to 75 years
PVR was measured by bladder ultrasonography at baseline, 1 year, and 2 years. Participants completed questionnaires about urinary symptoms and provided urine samples for microbiologic evaluation.
Of the 987 participants, 79% had a PVR <50 mL, 10% had a PVR of 50–99 mL, 6% had a PVR of 100–199 mL, and 5% had a PVR ≥200 mL at baseline. Of women with a PVR <50, 66% reported at least one urinary symptom at baseline. Of women with a PVR ≥200 mL, 27% reported no significant symptoms at baseline. In adjusted analyses using data from all study visits, women with a PVR ≥100 mL were more likely to report urinating >8 times during the day (OR=1.42, CI=1.07–1.87), and women with a PVR ≥200 mL were more likely to report weekly urgency incontinence (OR=1.50, CI=1.03–2.18), versus those with a PVR <50 mL. Elevated PVR was not associated with increased risk of stress incontinence, nocturnal frequency, or urinary tract infection in adjusted analyses. Forty-six percent of those with a PVR ≥200 mL and 63% of those with a PVR of 100–199 mL at baseline had a PVR <50 mL at 2 years.
Over 10% of ambulatory older women may have a PVR ≥100 mL associated with increased risk of some urinary symptoms, but many with elevated PVR are asymptomatic, and elevations in PVR frequently resolve within 2 years. Symptom-guided management of urinary symptoms may be more appropriate than PVR-guided management in this population.
Urinary incontinence is a common problem in middle-aged and older women, with up to a third of women aged 40 years and older and up to half of women aged 65 and older reporting weekly or more frequent incontinence.1, 2 Clinical guidelines from the Agency for Healthcare Research and Policy (formerly the Agency for Health Care Policy and Research, or AHCPR) and other professional organizations have recommended assessment of postvoid residual (PVR) bladder volume in all patients presenting with incontinence, to assess whether problems with bladder emptying may contribute to incontinence or complicate its treatment.3, 4 Nevertheless, while up to half of older men develop urinary symptoms as a result of prostatic enlargement,5 problems with bladder emptying are believed to be much less common in older women. As a result, some expert groups have questioned the utility of routinely measuring PVR in women, particularly in primary care settings where objective assessment of PVR by bladder catheterization or ultrasonography is not always easy to perform or readily available.6–8
Additionally, for both men and women, there is lack of consensus about what constitutes a significantly elevated PVR or at what point elevations in PVR start to contribute to urinary problems.6–8 According to AHCPR guidelines, a PVR of less than 50 mL is indicative of adequate bladder emptying, while a PVR of 200 or greater indicates inadequate emptying.3 However, these thresholds are based upon minimal data on the distribution of PVR volumes in the general population or evidence of associations between PVR and urinary symptoms. In the intermediate range or 50 to 200 mL, AHCPR guidelines offer no guidelines for interpretation other than for clinicians to exercise their own judgment.
With professional organizations calling for primary care clinicians to assume a greater role in diagnosing and treating incontinence,9, 10 there is a need for more data on normative values of PVR volume and relationships between PVR volumes and urinary outcomes. We examined the prevalence, natural history, and clinical significance of elevated PVR in a population-based cohort of ambulatory women aged 55 to 75 years, in order to inform decisions about using PVR measurements to evaluate urinary problems in this population.
Postvoid residual volume was measured in women aged 55 to 75 years in a 2-year prospective cohort study of risk factors for urinary tract infection (UTI) after menopause.11 All participants were enrollees in Group Health Cooperative, a group model health maintenance organization based in Washington state with approximately 450,000 members, who reported no natural menstrual period for at least one year. Women listed in the Group Health Cooperative’s diabetes registry were randomly sampled in the same age distribution as the main cohort to achieve a baseline prevalence of diabetes of approximately 20%, similar to the current prevalence of diabetes in older U.S. women.12 Exclusion criteria included residential nursing care, restriction to a wheelchair, dementia or severe psychiatric disorder, indwelling or intermittent urinary catheterization, end-stage renal disease requiring dialysis, active malignancy, acute cystitis in the preceding 90 days, or chronic antibiotic use. Between 1998 and 2002, 1,017 women were enrolled and followed for 2 years.
Postvoid residual volume was measured at baseline, 1 year, and 2 years, using a portable ultrasound device (BladderScan BVI 2500+, Verathon, Inc., Bothel, WA), a method that has been shown to correlate well with catheterized bladder volume measurement, especially for PVR volumes less than 200 mL.13–15 In preparation for their study visits, participants were told that they would be providing a urine sample at their study visit and to avoid emptying their bladder just before their visit. Measurement of PVR was performed at the end of each visit, after participants had finished reviewing and signing consent forms and completing all questionnaires about their medical history and urinary symptoms (approximately 30 to 45 minutes). Before measurement, women were instructed to go to the bathroom and empty their bladder, although women were not questioned about whether they had a normal urge to void. Afterwards, two separate bladder ultrasound measurements were immediately obtained with the participant lying in a supine position. All ultrasound measurements were obtained by a trained registered nurse who was responsible for PVR measurements throughout the study. According to the protocol, the study nurse uncovered the area between each participant’s umbilicus and pubic hairline, prepped and placed the scanhead at approximately 3 cm superior to the symphysis pubis so that it pointed toward the expected bladder location, and confirmed that the head of the icon on the scanhead was pointed toward the participant’s head. Ultrasound devices were set to the “FEMALE” setting unless participants indicated that they had undergone hysterectomy. After confirming that the participant’s bladder was centered in the crosshairs of the aiming display, the nurse obtained two separate bladderscan measurements using standard operating manual procedures.16 The average of the two measurements obtained at each visit was then recorded as the final PVR for that visit.
Urinary symptoms were assessed at baseline, 1 year, and 2 years using structured-item measures adapted from other large epidemiologic studies of urinary tract dysfunction.17–19 Women were first asked, “Have you had accidental leakage of urine during the past year?” followed by “During the past month about how often did this happen?” To characterize type of incontinence, women reporting at least monthly incontinence were asked, “Do you lose control of your urine because you feel the urge to urinate but cannot reach a bathroom in time?” for urgency-type and “Do you lose control of your urine when you laugh or cough or during physical activities?” for stress-type incontinence. Daytime and nocturnal frequency were assessed by asking “How many times do you have to urinate while you are awake during an average day (without a bladder infection)?” and “How many times do you usually awaken to urinate during the night (without a bladder infection)?”
Women were considered to have weekly urgency incontinence if they reported accidental leakage that occurred at least once a week and at least occasionally “with feeling the urge to urinate but not being able to reach a bathroom in time.” Stress incontinence was defined as accidental leakage occurring at least once a week and at least occasionally with “laughing or coughing or during physical activities.” Daytime frequency was defined as needing to urinate more than 8 times during an average day in the absence of a UTI. Nocturnal frequency was defined as needing to awaken more than once at night in the absence of a UTI. Urinary symptom categories were not mutually exclusive, in that participants could report more than one type of symptom.
Urinary tract infections were assessed throughout the 2-year study by asking women to contact research personnel at the onset of irritative voiding symptoms such as dysuria, urgency, or increased frequency. After confirming onset of symptoms, study staff instructed participants to collect home urine samples for microbiological evaluation using a Dipslide (Unipath, Ogdensburg, NY) home testing kit, supplied at enrollment along with instructions. Study staff also instructed symptomatic women to contact their primary physician for clinical evaluation, although participants were not systematically monitored for treatment or resolution of infection. Systematic surveillance of health plan clinical records was also performed; women with microbiological evidence of bacteriuria were interviewed by telephone to verify presence of urinary symptoms. Symptomatic, microbiologically-confirmed UTI was determined on the basis of a midstream, clean-catch urinary specimen yielding 105 or more colony forming units of a uropathogenic organism in the setting of at least two of the following symptoms: dysuria, urgency, or increased frequency.11
Other demographic and clinical variables were assessed by self-administered questionnaires. Prior UTI history was assessed by asking women, “Has a doctor or a nurse ever told you that you had a urinary tract infection, bladder infection, or cystitis?” Current diuretic use was assessed by asking women, “Do you take a diuretic (water pill) that makes you urinate more often?” followed by examples such as hydrochlorothiazide, furosemide, spironolactone, and triamterene. Use of other medications to control urination was assessed using a follow-up question, “Do you take another medication to help you control urination?” Overall health status was assessed through a standard single-item self-report measure in which participants rated their overall health as “excellent,” “very good,” “good,” “fair,” or “poor.”20, 21 Urinary tract surgery history was assessed by asking women, “Have you ever had surgery of the bladder or urinary system” and “Have you ever had surgery to help stop leakage of urine?”-- participants replying yes to either question were considered to have a history of urinary tract surgery. Body mass index was calculated at baseline from the health plan mammography surveillance database, based upon self-reported height and weight. Informed consent was obtained from all participants, and all study procedures were approved by the institutional review boards at both the University of Washington and Group Health Cooperative.
Given the potential effect of incontinence or overactive bladder medications on both PVR volume and urinary symptoms, statistical analyses were restricted to the 987 participants who successfully completed PVR measurement at baseline and did not report using a medication to control their urination other than diuretics at baseline (97.1% of the entire cohort). Data on other types of medications with potential anti-cholinergic properties were not available and therefore were not used to define the study population.
Demographic and clinical characteristics of participants, as well as baseline distribution and 2-year change in PVR volumes, were first examined using descriptive statistics. Next, generalized estimating equations (GEE) logistic regression models were developed to estimate the strength of associations between PVR elevation and urinary symptoms. Models used data from all visits (baseline, 1 year, and 2 years) and examined the relationship of PVR volume and urinary symptoms at each visit, while accounting for within-participant correlation of urinary outcomes across visits.22 Separate GEE models were developed to examine each urinary symptom (urgency incontinence, stress incontinence, daytime frequency, and nocturnal frequency), treated as four independent outcomes. The primary predictor in each model was PVR volume, categorized as 50 to 99, 100 to 199, or 200 mL or greater, versus PVR <50 mL as the reference group. All models were adjusted for age, race, body mass index, current use of diuretics, history of surgery to the urinary tract, and diabetes, as variables with the potential to affect both PVR and urinary symptoms. Multiple imputation was used in these models to address loss-to-follow-up as well as missing co-variate data. Twenty data sets were completed using the Markov chain Monte Carlo method under the assumption of multivariate normality. Summary effect estimates and standard errors were computed using standard methods for imputed data, implemented using the MI and MIanalyze procedures in SAS version 9.2.
Next, modified Cox proportional hazards models were developed to estimate independent associations of PVR volume at each study visit with risk of UTI at that visit or during the subsequent year, after confirming that there were no major violations of the proportional hazards assumption. Specifically, PVR elevations at baseline were examined in association with UTI risk between the baseline and 1-year visits, PVR elevations at the 1-year visit were examined in association with UTI risk between the 1-year and 2-year visits, and PVR elevations at the 2-year visit were examined in association with UTI status at the 2-year visit. In contrast to standard Cox models, participants were not removed from the analysis as soon as they had experienced a first outcome event, in order to ensure that information on recurrent UTIs was captured. Robust standard errors, which are estimated empirically from the data without making parametric assumptions about the correlation of the first and subsequent times to events, were used to account for potential clustering of events within participants.23 Models once again compared PVR volumes of 50 to 99, 100 to 199, and 200 or greater to PVR less than 50 mL as the reference group. These models were adjusted for age, race, body mass index, frequency of sexual activity, history of surgery to the urinary tract, and diabetes, as variables potentially related to both PVR and UTI risk. Multiple imputation was again used to address missing covariate data for variables such as body mass index. Statistical analyses were performed at the University of California San Francisco using SAS software, version 9.2 (SAS Institute, INC., Cary, NC), approved by the institutional review board of UCSF.
The mean (± SD) age of participants was 63.9 (± 6.3) years, and the majority were white (Table 1). Just under a third (N=312) met the study definition having urgency incontinence, and more than a third (N=351) met the definition of having stress incontinence. Approximately 30% (N=299) had daytime and 30% (N=305) had nocturnal frequency.
Postvoid residual measurements were within 10 mL of each other for 77.1% of visits, and within 25 mL of each other for 94.3% of visits. Of the 987 women in the analysis, 79.0% had a PVR less than 50, 9.6% had a PVR of 50 to 99, 6.8% had a PVR of 100 to 199, and 4.6% had a PVR of 200 mL or greater at baseline (Table 2). The majority of women in all PVR categories reported at least one urinary symptom at baseline. However, 33.5% of women with a PVR <50 mL, 29.5% of those with a PVR of 50 to 99, 29.9% of those with a PVR of 100 to 199, and 26.7% of women with a PVR of 200 or greater did not meet definitions for any symptom.
Of women with baseline PVR data, 78.1% (N=771) contributed PVR data at 2 years. Compared to women with no follow-up data, women who contributed 2-year PVR data were more likely to be older (mean [SD] age = 64.2 [6.3] versus 63.0 [6.0] years; P=.02), white (89.4% vs 81.7%; P<.01), college graduates (52.4% vs 40.3%; P<.01), in excellent or very good health (59.2% vs 42.9%; P<.01), and sexually active in the past month (33.3% vs 25.7%; P=.04), and were less likely to have a history of UTI in the past year (10.5% vs 17.2%; P<.01), diabetes (18.1% vs. 28.2%; P<.01), diuretic use (12.9% vs 19.1%; P=.02), obesity (26.3% vs 36.2%; P = .02), or smoking (43.2% vs 53.8%; P<.01). However, women contributing data at 2 years did not differ significantly from those lost to follow-up with regard to baseline urgency or stress incontinence, daytime or nocturnal frequency, or median PVR.
Over 2 years, the median (IQR) change in PVR was 0 (−22.5, 19.0) mL. Postvoid residual volumes changed by less than 25 mL in 54.3% of participants, but increased by 25 mL or more in 22.3%, and decreased by 25 mL or more in 23.4% (Table 3). Of women with a PVR of 200 mL or greater at baseline who contributed 2-year data on PVR (N=37), 45.9% (N=17) had a PVR less than 50 at 2 years. Of those with a PVR of 100 to 199 at baseline who contributed 2-year data (N=54), 63.0% (N=34) had a PVR less than 50 at 2 years.
In multivariable models adjusting for age, race, body mass index, diuretic use, urinary tract surgery, and diabetes, women with a PVR of 200 mL or greater were more likely to report urgency incontinence at any visit, compared to women with a PVR less than 50 (Table 4). Women with a PVR of 50 to 199 were also more likely to report daytime frequency compared to women with a PVR less than 50, although this was not observed for women with a PVR ≥200. Elevations in PVR were not associated with significantly increased odds of having stress incontinence or nocturnal frequency (Table 4).
A total of 138 symptomatic, microbiologically-confirmed UTIs were detected during the study, involving 97 participants. Postvoid residual volumes of 200 mL or greater were associated with increased 1-year risk of UTI compared to PVR of less than 50 in unadjusted models. After adjustment for potential confounders, however, the increased risk of UTI associated with a PVR of 200 or greater decreased and was no longer significant (Table 5).
In this prospective cohort of ambulatory women aged 55 to 75 years, more than 10% of women had a PVR of 100 mL or greater, and approximately 5% had a PVR of 200 or greater at baseline. Women with elevated PVR were more likely to report urinary symptoms such as daytime and nocturnal frequency compared to those with a PVR less than 50, but more than a quarter of women with a PVR of 200 or greater at baseline did not meet definitions for any urinary symptoms. Elevations in PVR were not associated with a significantly increased 1-year risk of UTI after adjustment for other demographic and clinical factors. Additionally, nearly half of women who had PVR volumes of 200 or greater at baseline were found to have a PVR of less than 50 after 2 years.
Clinical guidelines from the AHCPR and other organizations have recommended measurement of PVR in patients with urinary incontinence, and assessment of PVR has been evaluated as an indicator of quality of care in both institutional and community-based settings.24, 25 Nevertheless, there is uncertainty about the most appropriate way to use information about PVR to manage incontinence and other urinary symptoms, particularly in women. Among women who are candidates for anti-muscarinic therapy, some clinicians choose to withhold medication from those with elevated PVR, due to concern that treatment may further diminish bladder emptying, worsen urinary symptoms, or increase risk of UTI.7 Some clinicians also teach women to perform self-catheterization if they are found to have an elevated PVR in the setting of urinary symptoms or possible complications of urinary retention.6, 26 Additionally, in the primary care setting, many clinicians omit evaluation of PVR,27 while others refer all women with high PVR to a urologic or urogynecologic surgeon due to concern that a structural problem with the bladder, urethra, or pelvic structures may be interfering with bladder emptying.6, 28 Evidence to support these management strategies is lacking, however, and given the lack of data to guide determinations of “abnormal” PVR in women, routine PVR assessment may contribute to inconsistency in care without improving outcomes in this population.
These findings suggest that elevated PVR volumes of 100 mL or greater may not be as uncommon in ambulatory older women as previously thought. Prior research on PVR in women has tended to focus on small numbers of women drawn from referral, institutionalized, or rehabilitation populations; conversely, a few studies have examined PVR in women denying all urinary symptoms or risk factors for bladder dysfunction.29–34 In an English-language PubMed search using the terms to “postvoid residual,” “residual urine,” or “postvoid volume,” only 2 prior studies have provided data on PVR in older women that did not exclude or select participants on the basis of urinary symptoms.35, 36 In one study of 77 women from Western Copenhagen county, Denmark, all aged 75 years, the mean PVR by bladder ultrasound was 45 mL, with a 95th percentile of 168 mL.34 In another study of 169 women aged 60 years and older in Washtenaw county, Michigan, 16% of women were found to have a PVR of 50 mL or greater, and 9% had a PVR of 100 or greater.35 Only the latter study examined relationships between PVR and urinary symptoms, noting simply that the prevalence of any incontinence did not differ among participants with and without PVR greater than 50.
These findings also suggest that even when older women are found to have substantially elevated PVR, these elevations may resolve in a substantial proportion of women and do not necessarily indicate a persistent abnormality in bladder emptying. Variations in PVR measurements over time may reflect not only physiologic changes in bladder function, but also variability in patient effort or circumstances. Previous studies have documented variations in PVR in elderly men and women measured at different times during the day or over a period of weeks.37, 38 Additionally, clinical skill in measuring PVR volume by ultrasound is necessary, and variations in operator skill may result in less accurate PVR measurements. Combined with that research, our findings suggest that PVR measurements at a single visit – even when taken twice and obtained by clinically-trained personnel – may not be a reliable indicator of long-term bladder function in older ambulatory women.
While recognizing that elevations in PVR may contribute to urinary symptoms in older women, our findings tend to support symptom-guided rather than PVR-guided management of problems such as urgency incontinence, daytime frequency, or nocturnal frequency. Although elevations in PVR volume were associated with increased risk of urinary symptoms in this study, a substantial proportion of women with elevated PVR did not meet definitions for any symptoms, and approximately two thirds of women with a “normal” PVR of less than 50 mL experienced at least one symptom. Recently, some expert groups have suggested that it may be appropriate for clinicians to defer PVR measurement in women without major co-morbidities unless they show a paradoxical or adverse response to empiric therapy.6–8 Nevertheless, prospective interventional studies are needed to confirm the efficacy and safety of this strategy in women presenting in general practice.
Strengths of our study include the large sample size of older community-dwelling women who were not specifically selected on the basis of urinary symptoms, as well as use of longitudinal data to critically examine relationships between PVR and urinary symptoms over time. However, several limitations of this research should be noted. First, while the structured questionnaire measures distinguished between urgency and stress incontinence symptoms, they did not permit quantification of the number of urgency- or stress-type incontinence episodes per day. It is possible that a different pattern of associations with PVR would have emerged had this study used detailed voiding diaries to characterize women’s incontinence symptoms. Second, the absolute number of women with a PVR of 200 mL or greater in this study sample was small, which may have limited power to detect associations between PVR volumes in this range and urinary outcomes such as UTI. Nevertheless, prior published studies have included almost no women with a PVR of 200 or greater, and therefore this study provides some of the only data on the clinical significance of PVR volumes in this range.
Additionally, follow-up data on PVR volumes at 2 years were not available for 22.3% of women who underwent PVR measurement at baseline. If women with persistently elevated PVR were overrepresented among those lost, then this study may have underestimated the prevalence of persistent PVR elevation. Nevertheless, even assuming that all 8 women with a baseline PVR of 200 or greater who failed to provide 2-year follow-up data had a persistently elevated PVR, more than a third of women in this range would still have demonstrated resolution of their elevated PVR after 2 years.
Participants in this research were aged 55 to 75 years, which may limit the generalizability of these results to women older than 75 years. Within this cohort, however, there was no independent association between age and increased risk of PVR ≥50 mL at baseline (OR=0.92, 95%CI=0.77–1.10 per 5-year increase in age). Participants in this research were also predominantly white, college-educated, married, and in good health, and had a higher prevalence of estrogen use compared to the current general population. However, preliminary analyses also revealed no baseline associations between race, marital or education status, health status, or estrogen use and PVR >50 (P>.10 for all). Furthermore, prior research has not upheld an association between estrogen use and PVR in middle-aged or older women.30
In summary, this prospective study of PVR volume in ambulatory women aged 55 to 75 years suggests that elevations in PVR of 100 mL or greater may occur in over 10% of ambulatory women in this age group, and that PVR elevations in this range may be associated with increased risk of urinary symptoms such as daytime frequency and urgency incontinence. Nevertheless, given that elevations in PVR resolve in up to half of these women within 2 years and occur in a substantial proportion of women with no significant urinary symptoms, symptom-guided management of urinary symptoms may be more appropriate than PVR-guided management in many women.
Sources/research funding: This research was supported by NIH grants RR024130, T32AI07140, and RO1 DK43134, and by the resources and facilities at the Veterans Affairs Puget Sound Health Care System, Seattle, Washington. The views expressed in this paper are the authors and do not necessarily represent those of the NIH or the Department of Veterans Affairs.
Sponsor’s Role: The views expressed in this paper are those of the authors and do not represent those of the NIH, the Department of Veterans Affairs, or any other organization. No funders had any role in the design or conduct of this study; collection, management, analysis, and interpretation of the data; or preparation, review, or approval of the manuscript.
Previous presentation: This research was presented at the national scientific meeting of the American Geriatrics Society in Orlando, Florida, on May 13, 2010.
Conflicts of Interest:
|Elements of Conflicts||A.H.||J.B.||E.B.||E.M.||D.S.||L.W.||F.L.||E.V.||S.F.|
|Employment or Affiliation|
Author Contributions: Dr. Huang contributed significantly to the design of this study, analysis and interpretation of the data, drafting and critical revision of the article, and final approval of the version to be published. Drs. Brown, Moore, Walter, Lin, and Vittinghoff contributed significantly to analysis and interpretation of data, critical revision of the article, and final approval of the version to be published. Drs. Scholes, Boyko, and Fihn contributed significantly to the acquisition of the data, analysis and interpretation of data, critical revision of the article, and approval of the version to be published.
Explanation of conflicts of interest: Drs. Huang and Brown have received support for research related to urinary incontinence from Pfizer, Inc. via contracts with the University of California, San Francisco. No other authors have any potential conflicts of interest with regard to this research.