Variation in bowel and bladder continence across US spina bifida programs: A descriptive study

1. Introduction

Spina bifida (SB), a complex and multisystem birth defect resulting from abnormal formation of the neural tube during gestation, occurs in approximately 3–6 cases per 10,000 people in the United States [1, 2, 3]. Common primary and secondary conditions include skin break down, learning and developmental delays, orthopedic problems, sensory and motor impairment, hydrocephalus, Chiari malformation, as well as bowel and bladder dysfunction [4]. Perhaps more than any other factor, bowel and bladder impairments can result in profound social stigma and decreased self-esteem [5], as well as serious health complications including renal failure and even death [6, 7]. Unfortunately, neurogenic bowel and bladder are common in individuals with SB [4, 8], making it difficult for them to achieve continence despite advances in medical and surgical interventions.

Extant research provides widely discrepant estimates of bowel and/or bladder continence in SB ranging from 20–87% [9, 10, 11, 12, 13, 14, 15, 16, 17, 18]; a significant percentage of individuals experience “double incontinence” (i.e., both bowel and bladder) [16]. Discrepant findings in continence estimates are likely due to multiple factors. Of particular note are variations in patient phenotypes, sampling approaches, data collection methods, and definitions of continence. Further, many studies have limited their focus to participants from single clinics [19, 20], which does not allow evaluating outcome variations across clinical practices.

To overcome limitations of extant data on care, health status, and outcomes for individuals with SB (including continence), starting in 2008 the Centers for Disease Control and Prevention (CDC) began funding multidisciplinary SB clinics to implement the National Spina Bifida Patient Registry [21]. Using standardized variable definitions and processes of data collection, the NSBPR was designed to allow for a detailed description of the patient population attending SB-specific clinics and identification of processes of care that are associated with outcomes.

The purpose of this descriptive study was to measure bowel and bladder continence in patients with SB at multiple comprehensive, multidisciplinary SB clinics that are part of the NSBPR using a standardized questionnaire with a common and strict definition of continence and to determine the variation in continence prevalence across clinical programs. Describing findings from the NSBPR has the potential for contributing to a better understanding whether variations in continence previously published may be accounted for by use of different continence definitions and data collection methods across studies, or may reflect actual variations across clinic populations.

2. Patients and methods

3. Results

4. Discussion

To our knowledge, this multi-center study is the largest to investigate the prevalence of urinary and bowel continence in individuals with SB, and the first to examine variation across clinics in these variables. Using uniform definitions of “dry…during the day” and “no involuntary stool leakage … during the day,” only 40.8% and 43.0% of the entire cohort of 3252 individuals aged 5 years and older were reported as continent of urine and stool, respectively. These figures were lower when only those with myelomeningocele (37.6% and 39.6%) and those with bladder/bowel impairment (33.4% and 31.3%) were analyzed. Age, sex, race/ethnicity, SB type, functional level of lesion, insurance type, mobility status (bowel continence only), and education status (bowel continence only) were significantly and independently associated with continence.

These findings are sobering to those who care for patients with SB. All individuals included in this study received care at specialized, multi-disciplinary SB clinics and should have had access to the best and latest medical and surgical interventions to gain both urinary and bowel continence. Despite this, less than half of the entire cohort reported continence, and approximately one-third of those with known bladder and bowel impairment were continent. This is occurring in an era in which SB care has advanced beyond survival and has been focusing on quality of life improvement [5, 28, 29]. While it is currently unclear whether continence for 100% of patients will ever be achieved, these data serve as a continued charge to professionals caring for individuals with SB to strive toward identifying and implementing interventions that result in the best possible continence status for each individual patient.

Variation between clinic sites in the NSBPR was considerable, with prevalence of continence ranging 16.9–50.8% for urine and 9.5–49.1% for stool, even after controlling for demographic and condition-specific patient variables recorded as part of the registry. Clearly, our findings showed that less severe types of SB had better continence, but lesion level and mobility do not reliably predict bladder and bowel impairment [30]. Restriction of analysis to only those with bladder and bowel impairment still showed wide variation in continence rates, so selection bias based upon neurologic differences between different clinic populations may not explain continence variation across clinics. While socio-economic and demographic factors were independently associated with continence, clinic site variation in continence persisted when controlling for these factors. It is possible that other patient characteristics not tracked in the NSBPR may explain some of this variation, or that the analytic approach did not allow for detection of some associations because they are nonlinear. Socio-economic factors were not fully characterized, and parental/familial education levels and compliance with therapies were not assessed and could account for some site-specific differences. Whether inclusion of different patient, intervention, adherence, or clinic variables in analyses would eliminate variation by site is an important area of future research.

Observed variation in continence across clinics may contribute to an understanding of why previous research has documented such wide variability across studies. That is, sampling procedures in previous studies estimating continence rates have varied significantly (e.g., single or small number of clinics, community survey sampling). Our findings raise the possibility that recruitment from single or a small number of clinics may result in findings that are more reflective of local variables affecting continence rather than condition-based variables that are generalizable across the population of individuals with SB. Additional research is needed to further explore this possibility.

Bladder and bowel continence rates may be lower in this study than in previous research due to the stricter definition of continence. There is great variability in definitions of continence in the literature [15], ranging from one or more soiling or wetting accident per month [13, 31], to complete absence of urinary leakage or fecal accidents [19]. Some studies have failed to provide a definition of continence [15, 16], whereas others have emphasized distinctions between continence and being socially dry [11]. The definitions used in our study does not make allowance for small volume or infrequent incontinence that may be caused by occasional imperfect compliance with bladder and bowel regimens or extenuating circumstances and may not truly reflect the successes of clinic sites or specific behavioral, medical, or surgical interventions based on other definitions of continence (e.g., social continence). This is being addressed in future studies by quantifying and qualifying incontinence in more nuanced ways in version 2 of the NSBPR questionnaire, which was instituted after these data were collected. Additionally, continence rates in our study may not be reflective of the entire SB population as those with no or minimal problems may not attend specialized SB clinics.

Variability in continence across clinic sites may be partially attributable to data collection methods. Although the registry had a single definition of continence for bladder and for bowel, the method in which the questions were asked was not standardized. Ideally, information was taken from the medical records, but may have been gathered via face-to-face interview or by the individual completing a standardized questionnaire. How questions are asked and by whom may affect responses [32]. The existing SB literature is illustrative of this potential variability in continence rates with utilization of retrospective chart review [19], patient interview [13, 31], and questionnaire-based assessment methods [12]. On the other hand, variability in continence between clinic sites could be related to practice differences in medical and surgical interventions – in how frequently they are utilized, the types utilized, and the skill in which they are performed/implemented. Attempts to explain variability in continence rates by examining variations in interventions across clinics were not conducted for the current study, but are being examined.

Another important potential explanation of the variation in continence across clinics is possible selection or enrollment bias. Specifically, at the time of collection of data included in the current analysis, participating clinics were required to submit data on a minimum number of patients annually, but were not required to enroll all eligible participants. Whether differences in those enrolled versus those who were eligible but not enrolled accounts for variations in continence observed cannot be ruled out. Procedural updates to the NSBPR starting in 2014 require participating sites to provide basic demographic and condition-specific information on non-enrolled individuals without the inclusion protected health information, which may allow for control of any possible selection in future analyses.

Finally, errors in data acquisition and entry cannot be eliminated in a large study encompassing 20 sites. However, the NSBPR involves rigorous data quality checks and assurance procedures to maximize data accuracy [21].

5. Conclusion

These findings are instructive for all who care for individuals with SB in that urinary and bowel continence rates are relatively poor. These rates appear to vary significantly between leading SB clinics in the US and may reflect practice differences. Goals of the NSBPR included the identification of variations in care and development of best practices [21]; it is hoped that these efforts can be incorporated into bladder and bowel management of all SB individuals to improve care and quality of life.