- Email: [email protected]
What determines severity among patients with painful functional bowel disorders?
THE AMERICAN JOURNAL OF GASTROENTEROLOGY © 2000 by Am. Coll. of Gastroenterology Published by Elsevier Science Inc.
Vol. 95, No. 4, 2000 ISSN 0002-9270/00/$20.00 PII S0002-9270(00)00728-0
What Determines Severity Among Patients With Painful Functional Bowel Disorders? Douglas A. Drossman, M.D., William E. Whitehead, Ph.D., Brenda B. Toner, Ph.D., Nicholas Diamant, M.D., Yuming J.B. Hu, Ph.D., Shrikant I. Bangdiwala, Ph.D., and Huanguang Jia, Ph.D. UNC Center for Functional GI and Motility Disorders, Division of Digestive Diseases, University of North Carolina, Chapel Hill, North Carolina; and the Departments of Medicine and Psychiatry, University of Toronto, Toronto, Ontario, Canada
OBJECTIVE: For patients with painful functional bowel disorders (FBD), physicians frequently make diagnostic and treatment decisions based on the severity of the pain reported; patients with severe painful complaints may receive extensive diagnostic tests and treatments. Therefore, it would be important to determine what clinical factors contribute to the judgment of severity among patients with FBD. The aim of this study was to identify the psychosocial, behavioral, and physiological (visceral sensitivity) factors that predicted severity in patients with moderate to severe FBD. METHODS: Two hundred eleven female patients with moderate or severe FBD, as determined by the Functional Bowel Disorder Severity Index, entered a multicenter treatment trial at the University of North Carolina and the University of Toronto. Patients filled out diary cards and were given questionnaires and physiological testing (rectal sensitivity using barostat). Analysis of covariance and logistic regression adjusting for demographic factors were performed to determine which factors distinguished patients at study entry with moderate from those with severe FBD. RESULTS: Patients with severe FBD were characterized by greater depression and psychological distress, poorer physical functioning and health-related quality of life, more maladaptive coping strategies, and greater health care utilization. There was a trend for patients with severe FBD to have lower rectal sensation thresholds. Regression analysis indicated that severity was best predicted by behavioral features: poorer daily physical function, difficulties related to eating, more phone calls to the physician, and more days in bed for GI symptoms. CONCLUSIONS: We conclude that patient illness behaviors are best correlated with severity in FBD. The use of psychopharmacological agents (e.g., antidepressants) and psychological treatments to treat psychiatric comorbidity and to improve behavioral coping styles is recommended. Training to help medical physicians identify and respond to psychoThis work was presented at the Annual Meeting of the American Gastroenterological Association, May 16, 1999, Orlando, Florida.
social and behavioral features of these conditions is likely to improve patient satisfaction with their care and the clinical outcome. (Am J Gastroenterol 2000;95:974 –980. © 2000 by Am. Coll. of Gastroenterology)
INTRODUCTION The functional bowel disorders (FBD) are a subset of functional gastrointestinal (GI) disorders in which symptoms attributable to the mid or lower GI tract are not explained by structural or biochemical abnormalities (1). Because these disorders cannot be assessed through morphological or quantitative methods, physicians make clinical decisions based on their patients’ symptoms. This has led to the development of a symptom-based classification system for diagnosis in clinical care and research (2). Within the FBD classification, there are certain disorders (e.g., irritable bowel syndrome [IBS], chronic functional abdominal pain, and painful constipation) where pain is the predominant complaint. With these disorders, physicians will also make diagnostic and treatment decisions based on the severity of the patient’s report of pain. Therefore, a patient with IBS having mild discomfort may receive little or no diagnostic evaluation, and a recommendation for diet modification. However, a patient with the same diagnosis who reports severe and disabling pain may receive extensive diagnostic testing and treatment (e.g., surgical or psychiatric consultation, invasive imaging studies such as endoscopic retrograde cholangiography, use of antidepressants, or experimental medications) (3, 4). The factors contributing to the patient’s report of pain in FBD, and to the physician’s diagnostic and therapeutic decisionmaking, are poorly understood. The purpose of this study was to assess the relative contributions of certain hypothesized clinical factors (psychosocial, behavioral, and physiological) to the severity of painful FBD. A better understanding of the determinants of severity in these conditions should lead to the development of more effective diagnostic and therapeutic strategies that are based on documented predictive factors.
AJG – April, 2000
Severity in Patients With Painful FBD
MATERIALS AND METHODS Study Sample We studied female patients, aged 18 – 65 yr, enrolling between 1996 and 1998 into a National Institutes of Health multicenter treatment trial for FBD at the University of North Carolina at Chapel Hill and the University of Toronto. The study subjects were required to have painful functional bowel symptoms (categorized as IBS, painful functional constipation, chronic functional abdominal pain, or unspecified painful functional bowel disorder) (1) for at least 3 months, and to have symptoms for at least 2 days of the week. We used a previously validated Functional Bowel Disorder Severity Index (FBDSI) (5) to define severity in the patient groups. Similar to the method used to develop the Crohn’s Disease Activity Index (6), the physicians’ global rating of severity for their patients was used as the dependent variable in a regression analysis. The clinical variables that best predicted this global rating in the final model were then used to determine the FBDSI. The FBDSI ⫽ [current pain by visual analog scale (0 –100)] ⫹ [diagnosis of chronic functional abdominal pain 0 if absent; 106 if present] ⫹ [number of physician visits over 6 months ⫻ 11]. For this study, patients with mild symptom severity (FBDSI ⬍37) were excluded from participation in the study. The remaining patients were stratified either into moderate (FBDSI ⫽ 37–110) or severe (FBDSI ⬎110) groups. Comparisons were made at study entry between moderate and severe patient groups with regard to the factors listed below. Measurement— Clinical and Psychosocial Factors Before and during the baseline evaluation period (and before entering treatment), the study subjects were given diary cards, self-report questionnaires, and underwent physiological investigation. For this study, the following measures were assessed: A) AVERAGED DAILY PAIN INTENSITY. Patients kept a daily diary for 2 wks before the study visit. Daily pain intensity was assessed using a 100-mm visual analog scale (0 ⫽ no pain, 100 ⫽ very severe), and average score for the 14 days was used in this analysis. B) PSYCHOSOCIAL QUESTIONNAIRES. At the time of the initial study visit, patients received the following standardized psychosocial questionnaires: 1. SCL-90 (7). The Global Symptom Index (GSI) was used as a state measure of psychological distress; 2. Beck Depression Inventory (BDI) (8). The BDI assessed symptoms of depression; 3. NEO Inventory (9). The Neuroticism scale of the NEO Inventory was used to assess personality dysfunction relating to resiliency in the face of stress; 4. Sickness Impact Profile (SIP) (10). The SIP is a generic health status instrument that focuses on daily function.
5.
6.
7.
8.
9.
975
We evaluated overall daily function (Overall Scale), in addition to the following subscales: Physical Function, Sleep and Rest, Home Management, Work, Recreation and Pastime, and Eating. Psychosocial function was not evaluated because it was too closely correlated with the other psychosocial measures. Social Support Questionnaire (SSQ-6) (11). The SSQ-6 measures social support using two subscales: Number of Supportive People, and Degree of Satisfaction with the support. Dysfunctional Attitudes Scale (DAS). The DAS is used to assess dysfunctional attitudes that may relate to symptom reporting, for example: “What other people think of me is very important.” Coping Strategies Questionnaire. (CSQ) (12). The CSQ is a coping measure frequently used for patients with chronic pain. We used specifically the Catastrophizing Scale of the CSQ and also two questions relating to subjects’ ability to control and decrease their symptoms. These items are predictive of poor health status in patients with functional GI disorders (13). IBS-QOL (14). The IBS-QOL is a well-validated condition-specific measure of quality of life for patients with IBS and other functional bowel disorders. Sexual/physical Abuse History (15). We used a confidential survey questionnaire from previous studies (15) to assess abuse history.
C) HEALTH CARE UTILIZATION/BEHAVIORAL MEASURES. We included the following: 1. Number of days in bed for GI symptoms over a 3-month period; 2. Number of overnight admissions (for all and just GI symptoms) over a 3-month period; 3. Number of total nightly stays (for all and just GI symptoms) over a 3-month period; 4. Number of total procedures (surgical, endoscopic) over the previous 3 months; 5. Number of M.D. visits (for all and just GI symptoms) over a 3-month period; 6. Number of times phoned GI physician over a 3-month period; 7. Number of hospitalizations in the last 2 yr; 8. Number of lifetime surgeries. D) STANDARDIZED RECTAL SENSATION THRESHOLDS. The Synectics Visceral Stimulator (MedTronic Synectics Medical, Shoreview, MN) was used to measure muscle tone and sensory thresholds by the barostat technique. Standardized assessment (16) of pressure and volume was obtained for: 1. 2. 3. 4.
Urge threshold (ascending method of limits); 1st pain report (ascending method of limits); Pain threshold (ascending method of limits; Pain threshold (tracking technique).
976
Drossman et al.
Statistical Analysis All data were entered into a data set using SAS Full Screen Editor, and analyzed using SAS (SAS, Cary, NC). Comparisons of demographic and clinical data were performed between study sites to assure comparability of the data for later analyses where data from these groups were combined. The UNC and Toronto data were used to compare moderate and severe patient groups for all except the physiology data, which was obtained from UNC only. We first performed GLM ANCOVA (adjusting for age, race, and education) for continuous variables, and 2 tests for categorical data to assess the relationship between the various factors with severity group status. After this, those factors that significantly discriminated between moderate and severe groups were entered as independent variables into a multiple logistic regression model, using severity group status (moderate vs severe) as the dependent variable. Because some psychosocial variables might be linearly associated, we performed collinearity testing to ascertain the appropriateness of the independent variables that were entered into the final logistic model. Although abdominal pain (VAS) score and number of MD visits/3 months discriminated moderate from severe groups, they were not included in the model, because these two variables are used to create the FBDSI. In addition, the SIP Overall Scale and the SIP Psychosocial subscale were not included because of collinearity effects, using condition indices ⬎15. The results from the multiple logistic regression are presented in terms of p values, adjusted odds ratios (OR), and 95% confidence intervals (CI).
RESULTS Study Population There were 211 female patients with FBD (108 Chapel Hill, 103 Toronto) who enrolled during the study period, of whom 84% were white (79% Chapel Hill, 89% Toronto; p ⬍ 0.04). The mean age was 39.3 ⫾ 12.1 SD, with 14.6 ⫾ 2.9 yr educational attainment. There were 144 patients (79 Chapel Hill, 65 Toronto) with moderate FBD and 67 patients (29 Chapel Hill, 38 Toronto) with severe FBD. The physicians identified 83.3% to have IBS, 3.3% painful functional constipation, 6.7% chronic functional abdominal pain, and 4.3% unspecified functional bowel disorder. There were no statistical differences between Chapel Hill and Toronto with regard to age, education, pain scores, and psychosocial and utilization/behavioral variables. Therefore, data from the two sites were combined in the later analyses. Comparison Between Moderate and Severe Groups DEMOGRAPHIC FACTORS. Patients with severe FBD had 1 yr less educational attainment (Moderate, 14.9 ⫾3.0 yr vs Severe, 13.9 ⫹ 2.5 yr; p ⬍ 0.01), and there were no
AJG – Vol. 95, No. 4, 2000
differences between groups with regard to race and age. Subsequent analyses were performed adjusting for age, race, and educational attainment. AVERAGED DAILY PAIN. Patients with severe FBD reported significantly greater pain (Severe, 51.0 ⫾ 2.5 vs Moderate, 35.1 ⫾ 1.7; p ⬍ 0.0001) on the 14-day averaged Visual Analog Scale score (0 ⫽ no pain, 100 ⫽ very severe pain). This pain measure was not used in further analyses, as pain is a defining variable for the FBDSI. PSYCHOSOCIAL SCORES. Table 1 compares the moderate and severe groups on the psychosocial scores, adjusting for age, race, and education. Patients with severe FBD reported significantly higher scores on the the Beck Depression Inventory (p ⬍ 0.005), the Overall Scale of the Sickness Impact Profile (p ⬍ 0.0001), and most subscales of the SIP (p ⬍ 0.01 to p ⬍ 0.0001). Patients with severe FBD also reported higher scores on the Catastrophizing Scale of the Coping Strategies Questionnaire (p ⬍ 0.0001), as well as less perceived ability to control (p ⬍ 0.05) or decrease (p ⬍ 0.01) symptoms. Finally, the severe FBD group showed significantly poorer quality of life on the Overall score on the IBS-QOL (p ⬍ 0.0003). There were no statistically significant differences with regard to SCL-90R GSI (trend at p ⬍ 0.07), neuroticism, social support, or dysfunctional attitudes. We also compared the frequency of physical and sexual abuse reported by questionnaire. There was a higher frequency of all types of abuse by the moderate group (57% vs 40%; p ⬍ 0.02). When evaluating the frequency of rape, the difference between groups was no longer statistically significant (22.2% vs 13%; p ⫽ 0.13). UTILIZATION AND BEHAVIORAL SCORES. Table 2 compares the moderate and severe groups with regard to the health care utilization and behavioral data, adjusting for age, race, and education. Patients with severe FBD reported more days in bed for GI symptoms (p ⬍ 0.0001), more overnight admissions for GI problems ⬎3 months (p ⬍ 0.03), more physician visits for GI and other symptoms ⬎3 months (p ⬍ 0.0001), more phone calls for GI symptoms ⬎3 months (p ⬍ 0.0001), and more hospitalizations in the previous 2 yr (0.0001). There was a trend toward more nightly stays in the hospital (p ⬍ 0.08) and more lifetime surgeries (p ⬍ 0.10). There was no difference between groups for the number of endoscopic and surgical procedure over the previous 3 months. RECTAL SENSATION THRESHOLDS. Table 3 compares the moderate and severe groups with regard to rectal thresholds to balloon distension using the barostat, adjusting for age, race, and education. There was a trend toward lower thresholds for the severe FBD group, but significance was achieved for only one of eight variables (volume at 1st pain report using the ascending method of limits; p ⬍ 0.05).
AJG – April, 2000
Severity in Patients With Painful FBD
977
Table 1. Comparison Between Moderate and Severe FBD: Psychosocial Tests Psychosocial Scale
Moderate FBD Mean (S.E.)
Severe FBD Mean (S.E.)
p⬍
SCL-90 (GSI) Beck Depression Index NEO (Neuroticism) SIP Overall Score SIP Physical SIP Sleep/Rest SIP Home Management SIP Work SIP Recreation SIP Eating Social Support Questionnaire SSQ # People SSQ Satisfaction Dysfunctional Attitudes Perfectionism Approval by others Coping Strategies Questionnaire Catastrophizing Control of symptoms Decrease of symptoms IBS-QOL Overall
57.6 (3.5) 9.3 (0.7) 89.2 (2.0) 5.5 (0.5) 2.2 (0.3) 10.2 (1.0) 5.5 (0.9) 9.8 (1.9) 13.1 (1.3) 2.0 (0.4)
69.3 (5.2) 12.5 (1.0) 91.4 (3.0) 9.3 (0.8) 5.4 (0.5) 13.6 (1.6) 10.8 (1.3) 18.5 (2.9) 18.7 (1.9) 5.1 (0.5)
0.07 0.005 0.54 0.0001 0.0001 0.07 0.001 0.01 0.001 0.0001
21.1 (1.0) 30.6 (0.5)
19.1 (1.5) 30.4 (0.7)
0.28 0.84
83.3 (1.3) 44.5 (0.8)
86.2 (1.9) 44.0 (1.2)
0.20 0.75
8.2 (0.6) 2.6 (0.1) 2.4 (0.1) 69.3 (1.6)
12.9 (0.8) 2.2 (0.2) 1.9 (0.2) 58.8 (2.4)
0.0001 0.05 0.01 0.0003
FBD ⫽ functional bowel disorders; SIP ⫽ Sickness Impact Profile. For explanation of psychosocial scales used, see Materials and Methods section.
Multiple Logistic Regression
DISCUSSION
To determine the relative importance of the factors that best predict severity, a logistic regression model was fitted with Severity (Moderate vs Severe) as the dependent variable, and the significant variables from the previous ANCOVAs (Tables 1–3) were considered in a stepwise fashion as independent predictor variables. Table 4 shows the results from the final logistic regression. Of the 19 variables entered into the model, only four—Physical Dysfunction (SIP) (p ⬍ 0.01), Eating Dysfunction (SIP) (p ⬍ 0.027), days in bed for GI symptoms (p ⬍ 0.005), and number of times phoned M.D. for GI symptoms (p ⬍ 0.009) significantly predicted severity. We also ran a separate regression analysis including the visceral sensitivity data as predictor variables for the UNC data set and the visceral sensitivity variables were not significant predictors (data not reported).
For the irritable bowel syndrome (IBS) and other functional bowel disorders (FBD), diagnosis and treatment usually depend on the predominant symptom pattern. Thus, patients with predominant diarrhea will have different diagnostic studies and treatments than patients with predominant constipation or abdominal bloating (4). This has led to the development and use of a symptom-based diagnostic classification system, the Rome criteria. (2) However, when pain is a dominant feature of the condition, physicians also make these decisions based on another clinical dimension, the severity of the pain reported (3, 4, 17). Reports of pain that is mild in intensity may lead to little or no diagnostic evaluation, and treatment may include simple reassurance or a recommendation for diet and lifestyle change. But reports of severe pain may lead to extensive diagnostic eval-
Table 2. Comparison Between Moderate and Severe FBD: Health Care Utilization and Behaviors Health Care/Behaviors
Moderate FBD Mean (S.E.)
Severe FBD Mean (S.E.)
p⬍
Days in bed for GI/3 months Overnight admissions/3 months Overnight admissions GI/3 months Total nightly stays/3 months Total nightly stays GI/3 months Total procedures Total MD visits/3 months Total MD visits GI/3 months Times phoned GI MD/3 months Hospitalizations/2 yr Lifetime surgeries
2.2 (0.9) 0.02 (0.02) 0.002 (0.01) 0.02 (0.2) 0.01 (0.08) 0.08 (0.03) 2.1 (0.40) 1.1 (0.3) 0.4 (0.2) 0.2 (0.09) 2.4 (0.2)
11.8 (1.4) 0.06 (0.03) 0.06 (0.02) 0.5 (0.2) 0.03 (0.1) 0.05 (0.04) 6.9 (0.5) 4.8 (0.4) 1.6 (0.2) 0.9 (0.1) 3.1 (0.4)
0.0001 0.35 0.03 0.08 0.08 0.62 0.0001 0.0001 0.0001 0.0001 0.10
FBD ⫽ functional bowel disorders.
978
Drossman et al.
AJG – Vol. 95, No. 4, 2000
Table 3. Comparison Between Moderate and Severe FBD: Rectal Sensitivity Rectal Sensitivity Parameter Pain threshold Pressure Volume First pain report Pressure Volume Urge threshold Pressure Volume Pain threshold (tracking) Pressure Volume
Moderate FBD Mean (S.E.)
Severe FBD Mean (S.E.)
p⬍
30.0 (1.3) 164.6 (11.2)
27.6 (2.2) 125.9 (18.6)
0.36 0.08
32.6 (3.1) 170.7 (11.1)
23.8 (5.2) 127.9 (18.5)
0.16 0.05
27.0 (1.9) 151.9 (10.7)
22.0 (3.1) 119.3 (17.7)
0.17 0.12
29.9 (4.0) 195.6 (12.4)
22.9 (6.6) 149.1 (20.5)
0.38 0.06
FBD ⫽ functional bowel disorders.
uation (e.g., colonoscopy, computed tomography [CT] scan or endoscopic retrograde cholangiopancreatography [ERCP] and biliary manometry, or surgical or psychiatric consultation), and treatments (e.g., prescription of antidepressants, or employment of experimental medications and alternative treatments). Recent studies support an important role for psychosocial factors in influencing pain severity among patients with functional GI disorders. These factors may amplify the pain experience and/or increase illness-related behaviors that communicate the pain to others (4, 18). Although intestinal dysmotility contributes to many functional GI symptoms (e.g., diarrhea, constipation, vomiting), there is little evidence to show that it contributes to chronic GI pain. However, there is an association between visceral hypersensitivity and pain reports (19), which may be mediated at the level of the afferent receptor, the spinal cord, or higher centers, and current opinion favors a strong contribution from the central nervous system (CNS) (20, 21). In fact, brain-imaging studies implicate a regulatory role for the neocortex (e.g., anterior cingulate gyrus, prefrontal cortex) in the affective experience of visceral pain (22, 23). Finally, psychosocial disturbance may influence the development of functional GI symptoms after an acute enteric infection (24). This may occur through stress-mediated effects both on motility and also on inflammation via the HPA immune/ cytokine axis (25). Inflammation (possibly resulting from CNS activation of peripheral cytokines) can lead to intestiTable 4. Stepwise Logistic Regression: Final Model Predictor Variable*
Odds Ratio
95% CI
p⬍
SIP Physical Subscale SIP Eating Days in bed for GI/3 months # Phone calls to MD/3 months
0.87 0.90 0.92 0.67
(0.78–0.97) (0.82–0.99) (0.86–0.97) (0.48–0.88)
0.013 0.027 0.005 0.009
*The following variables were also entered into the stepwise logistic regression, but were eliminated from the final model: Age, race, education, Beck Depression Inventory, SCL-90 Global Severity, Catastrophizing Scale of CSQ, perceived ability to decrease or control symptoms (CSQ), IBS-QOL Overall Score, SIP Sleep and Rest, SIP Home Management, SIP Work, SIP Recreation, Days in bed for GI symptoms (3 months), # Hospitalization (2 yr), # Lifetime surgeries.
nal dysmotility and visceral sensitization, even after the acute infection resolves (26). The aim of the current study was to determine the factors that contribute to severity among patients with painful functional bowel disorders (IBS, painful functional constipation, and chronic functional abdominal pain). Our hypothesis was that psychosocial, and behavioral, more than physiological (visceral hypersensitivity) factors would predict severity, as determined by the Functional Bowel Disorder Severity Index (FBDSI) (5). As predicted, we found that most of the psychosocial and behavioral variables significantly discriminated these two patients groups, and visceral hypersensitivity had little or no effect. Furthermore, the logistic regression indicated that severity was best associated with patient behaviors rather than with psychosocial factors: their daily physical functional status and eating capability, the number of days spent in bed, and the number of phone calls to physicians because of GI illness. Why might these behaviors be such strong predictors of severity? We believe that the physician’s assessment of severity in FBD is influenced primarily by how the patient behaves in response to the pain (27). As previously shown with the development of the FBDSI, the physician’s global rating of pain was best determined by the severity and chronicity of the patient’s report of pain, and the number of physician visits. Furthermore, in the current study, the severe group was characterized by additional patient behaviors: restricting or modifying the diet because of pain (SIPEating subscale), frequently phoning the physician (number of phone calls/3 months), cutting back on daily physical activities (SIP-Physical subscale), and taking to bed (number of days in bed/3 months for GI problems). Although the physician perceives these behaviors as reflective of severe physical symptoms, a dilemma ensues, because, in many cases, diagnostic and therapeutic efforts directed toward the symptoms fail. This dilemma may result from how physicians understand and respond to patients with chronic pain behaviors. When applying the traditional biomedical model, physicians will search for a specific medical source for the pain and will employ medical treatments, expecting symptom relief (28). However, this expectation is not usually met: expensive diagnostic evaluations do not disclose other conditions (27), and treatments directed at the gut (e.g., antispasmodics, antidiarrheal drugs, etc.) are often unsuccessful (17). Furthermore, the underlying psychosocial factors contributing to the patient’s pain experience and behavior, as identified in this study, are usually ignored, minimized, or relegated to being secondary phenomena. As shown in one study of medical outpatients with IBS (29), physicians could more accurately identify the details of the patients’ somatic complaints than they could patients’ symptom-related cognitions, behavior, and anxieties, all of which are important components of the experience of pain. The result is that many patients with painful functional GI symptoms become dissatisfied with the care they receive,
AJG – April, 2000
and may seek the advice of other physicians or obtain alternative treatments (30). In turn, the physicians remain perplexed by the patients’ continued complaints despite no evidence for “disease.” They are frustrated in their unsuccessful efforts, may assume that the patient cannot be helped, and may develop rejecting or stigmatizing attitudes and behaviors. Ultimately this leads to a deterioration in the physician-patient relationship, and the perpetuation of a vicious cycle of refractoriness and help-seeking. For these reasons, we conclude that the biomedical model is neither satisfactory nor sufficient for patients with FBD. We would recommend a biopsychosocial approach in which the symptoms are understood in terms of the biological and psychosocial determinants (28). The findings of this study have important implications with regard to the application of this model in the care of patients with painful FBD. First, efforts can be made to reduce the pain (via central analgesia) and possibly psychiatric comorbidity (e.g., depression, panic) through the use of antidepressants. Second, efforts can also be made to modify pain-related behaviors, and increase cognitive coping strategies through referral to a mental health professional skilled in psychological techniques (e.g., interpersonal psychotherapy or cognitive therapy), which have been shown effective. Preliminary studies suggest that both antidepressants and psychological treatments improve the clinical outcome of patients with functional GI disorders, but further well-designed studies of large patient samples are needed. Finally, there is a compelling need to educate physicians on how to elicit relevant psychosocial data and integrate this information into a comprehensive, biopsychosocial plan of care. The physician can achieve this by (31) demonstrating a genuine interest in the patient’s beliefs and expectations relating to the condition; using patient-centered communication strategies (32); eliciting the patient’s concerns and offering appropriate reassurance; understanding and seeking to modify any somatic misattributions (e.g., that the patient has cancer); making educated decisions about the use of psychopharmacologic agents; and, when appropriate, arranging mental health referral concurrent with the medical care. It was recently shown that diagnostic investigations among patients with IBS did not improve patient satisfaction. However, when the physicians correctly identified and addressed the patients’ psychological attributions related to their IBS, their satisfaction with the care increased, and level of anxiety, and maladaptive cognitions diminished (33). Recent recommendations have been made by the American Gastroenterological Association to provide supervised time for clinical and psychosocial skill training for fellows seeing patients with functional GI disorders (34). There are three features relating to the study design or analysis that deserve further comment. First, the study involved patients with moderate to severe FBD seen mostly in referral (medical center) practices. Therefore, the conclusions drawn from these data may not apply to the majority of patients seen in primary care practice or to patients with
Severity in Patients With Painful FBD
979
milder or nonpainful symptoms. Patients with mild FBD were excluded by the protocol to understand this more difficult-to-treat group of patients. Second, we may have overlooked other important predictor variables that might contribute to severity. These might involve additional psychosocial (e.g., psychiatric or personality trait) or medical (e.g., stool frequency, vomiting, etc.) variables. Nevertheless, the available data compel us to conclude that psychosocial and behavioral factors are strong determinants of severity. Finally, in comparing differences between moderate and severe groups we did not adjust for multiple comparisons. Our preliminary (ANCOVA) analyses were descriptive in nature, and were also used to identify meaningful variables for the logistic regression.
CONCLUSIONS We conclude that patient illness behaviors are the best predictors of severity in FBD. The use of psychopharmacological agents (e.g., antidepressants) and psychological treatments may help reduce patient distress through treatment of psychological comorbidity and improvement in coping styles. In addition, training to help medical physicians identify and respond to psychosocial and behavioral features is likely to improve patients’ satisfaction with the care and the clinical outcome.
ACKNOWLEDGMENTS This study was supported by NIH Grants #RO1DK49334 and KO5MH00133, UNC Center for Functional GI and Motility Disorders, the UNC Center for Gastrointestinal Biology and Disease, and the UNC General Clinical Research Center. We thank Udi Ringel, M.D., for his review of this article. Reprint requests and correspondence: Douglas A. Drossman, M.D., Division of Digestive Diseases, Department of Medicine, University of North Carolina, 726 Burnett-Womack Building, CB #7080, Chapel Hill, NC 27599-7080. Received Apr. 29, 1999; accepted Sep. 7, 1999.
REFERENCES 1. Thompson WG. C. Functional bowel disorders and D. Functional abdominal pain. In: Drossman DA, Richter JE, Talley NJ, et al., eds. Functional gastrointestinal disorders: Diagnosis, pathophysiology and treatment, 1st ed. McLean, VA: Degnon Associates, 1994:115–73. 2. Drossman DA, Richter JE, Talley NJ, et al. Functional gastrointestinal disorders: Diagnosis, pathophysiology and teatment, 1st ed. McLean, VA: Degnon Associates, 1994:1–370. 3. Drossman DA, Thompson WG. The irritable bowel syndrome: Review and a graduated, multicomponent treatment approach. Ann Intern Med 1992;116:1009 –16. 4. Drossman DA, Whitehead WE, Camilleri M. Irritable bowel syndrome: A technical review for practice guideline development. Gastroenterology 1997;112:2120 –37.
980
Drossman et al.
5. Drossman DA, Li Z, Toner BB, et al. Functional bowel disorders. A multicenter comparison of health status, and development of illness severity index. Dig Dis Sci 1995;40:986 –95. 6. Best WR, Becktel JM, Singleton JW, et al. Development of a Crohn’s disease activity index. National Cooperative Crohn’s disease study. Gastroenterology 1976;70:439 – 44. 7. Derogatis LR. SCL-90-R: Administration, scoring, and procedures manual II —For the r(evised) version, 1st ed. Towson, MD: Clinical Psychometric Research, 1983. 8. Beck AT, Ward CH, Mendelson M, et al. An inventory for measuring depression. Arch Gen Psychiatry 1961;4:561–71. 9. Costa PT Jr, McCrae RR. The NEO personality inventory manual. Odessa, FL: Psychological Assessment Resources, 1985. 10. Bergner M, Bobbitt RA, Carter WB. The sickness impact profile: Development and final revision of a health status measure. Med Care 1981;19:787– 805. 11. Sarason IG, Sarason BR, Shearin EN, et al. A brief measure of social support: Practical and theoretical implications. J Soc Personal Relationships 1987;4:497–510. 12. Rosenstiel AK, Keefe FJ. The use of coping strategies in chronic low back pain patients: Relationship to patient characteristics and current adjustment. Pain 1983;17:33– 40. 13. Drossman DA, Li Z, Leserman J, et al. Effects of coping on health outcome among female patients with gastrointestinal disorders. Psychosom Med 2000 (in press). 14. Patrick DL, Drossman DA, Frederick IO, et al. Quality of life in persons with irritable bowel syndrome: Development of a new measure. Dig Dis Sci 1998;43:400 –11. 15. Drossman DA, Leserman J, Nachman G, et al. Sexual and physical abuse in women with functional or organic gastrointestinal disorders. Ann Intern Med 1990;113:828 –33. 16. Whitehead WE, Delvaux M, Working Team. Standardization of procedures for testing smooth muscle tone and sensory thresholds in the gastrointestinal tract. Dig Dis Sci 1994;42: 223– 41. 17. Drossman DA. Diagnosing and treating patients with refractory functional gastrointestinal disorders. Ann Intern Med 1995;123:688 –97. 18. Drossman DA, Creed FH, Olden KW, et al. Psychosocial aspects of the functional gastrointestinal disorders. Gut 1999; 45:II25–II30. 19. Mertz H, Naliboff B, Munakata J, et al. Altered rectal perception is a biological marker of patients with irritable bowel syndrome. Gastroenterology 1995;109:40 –52.
AJG – Vol. 95, No. 4, 2000
20. Mayer EA, Gebhart GF. Basic and clinical aspects of visceral hyperalgesia. Gastroenterology 1994;107:271–93. 21. Whitehead WE, Palsson OS. Is rectal pain sensitivity a biological marker for irritable bowel sydrome: Psychological influences on pain perception. Gastroenterology 1998;115:1– 10. 22. Silverman DHS, Munakata JA, Ennes H, et al. Regional cerebral activity in normal and pathologic perception of visceral pain. Gastroenterology 1997;112:64 –72. 23. Aziz Q, Thompson DG. Brain-gut axis in health and disease. Gastroenterology 1998;114:559 –78. 24. Gwee KA, Leong YL, Graham C, et al. The role of psychological and biological factors in post-infective gut dysfunction. Gut 1999;44:400 – 6. 25. Drossman DA. Mind over matter in the postinfective irritable bowel. Gut 1999;44:306 –7. 26. Collins SM. The immunomodulation of enteric neuromuscular function: Implications for motility and inflammatory disorders. Gastroenterology 1996;111:1683–99. 27. Drossman DA. Chronic functional abdominal pain. Am J Gastroenterol 1996;91:2270 – 81. 28. Drossman DA. Presidential address: Gastrointestinal illness and biopsychosocial model. Psychosom Med 1998;60:258 – 67. 29. Van Dulmen AM, Fennis JFM, Mokkink HGA, et al. Doctors’ perception of patients’ cognitions and complaints in irritable bowel syndrome at an out-patient clinic. J Psychosom Res 1994;38:581–90. 30. Sutherland LR, Verhoef MJ. Why do patients seek a second opinion or alternative medicine? J Clin Gastroenterol 1994;3: 194 –7. 31. Drossman DA. Psychosocial factors in the care of patients with gastrointestinal disorders. In: Yamada T, ed. Textbook of gastroenterology, 3rd ed. Philadelphia: Lippincott-Raven, 1999:638 –59. 32. Drossman DA. Psychosocial sound bites. Exercises in the patient-doctor relationship. Am J Gastroenterol 1997;92: 1418 –23. 33. Van Dulmen AM, Fennis JFM, Mokkink HGA, et al. Doctordependent changes in complaint-related cognitions and anxiety during medical consultations in functional abdominal complaints. Psychol Med 1995;25:1011– 8. 34. Ouyang A, Camilleri M, Drossman DA, et al. Task force on training in motility, diverticular disease, and functional illness. Gastroenterology 1996;110:1274 – 6.
Vol. 95, No. 4, 2000 ISSN 0002-9270/00/$20.00 PII S0002-9270(00)00728-0
What Determines Severity Among Patients With Painful Functional Bowel Disorders? Douglas A. Drossman, M.D., William E. Whitehead, Ph.D., Brenda B. Toner, Ph.D., Nicholas Diamant, M.D., Yuming J.B. Hu, Ph.D., Shrikant I. Bangdiwala, Ph.D., and Huanguang Jia, Ph.D. UNC Center for Functional GI and Motility Disorders, Division of Digestive Diseases, University of North Carolina, Chapel Hill, North Carolina; and the Departments of Medicine and Psychiatry, University of Toronto, Toronto, Ontario, Canada
OBJECTIVE: For patients with painful functional bowel disorders (FBD), physicians frequently make diagnostic and treatment decisions based on the severity of the pain reported; patients with severe painful complaints may receive extensive diagnostic tests and treatments. Therefore, it would be important to determine what clinical factors contribute to the judgment of severity among patients with FBD. The aim of this study was to identify the psychosocial, behavioral, and physiological (visceral sensitivity) factors that predicted severity in patients with moderate to severe FBD. METHODS: Two hundred eleven female patients with moderate or severe FBD, as determined by the Functional Bowel Disorder Severity Index, entered a multicenter treatment trial at the University of North Carolina and the University of Toronto. Patients filled out diary cards and were given questionnaires and physiological testing (rectal sensitivity using barostat). Analysis of covariance and logistic regression adjusting for demographic factors were performed to determine which factors distinguished patients at study entry with moderate from those with severe FBD. RESULTS: Patients with severe FBD were characterized by greater depression and psychological distress, poorer physical functioning and health-related quality of life, more maladaptive coping strategies, and greater health care utilization. There was a trend for patients with severe FBD to have lower rectal sensation thresholds. Regression analysis indicated that severity was best predicted by behavioral features: poorer daily physical function, difficulties related to eating, more phone calls to the physician, and more days in bed for GI symptoms. CONCLUSIONS: We conclude that patient illness behaviors are best correlated with severity in FBD. The use of psychopharmacological agents (e.g., antidepressants) and psychological treatments to treat psychiatric comorbidity and to improve behavioral coping styles is recommended. Training to help medical physicians identify and respond to psychoThis work was presented at the Annual Meeting of the American Gastroenterological Association, May 16, 1999, Orlando, Florida.
social and behavioral features of these conditions is likely to improve patient satisfaction with their care and the clinical outcome. (Am J Gastroenterol 2000;95:974 –980. © 2000 by Am. Coll. of Gastroenterology)
INTRODUCTION The functional bowel disorders (FBD) are a subset of functional gastrointestinal (GI) disorders in which symptoms attributable to the mid or lower GI tract are not explained by structural or biochemical abnormalities (1). Because these disorders cannot be assessed through morphological or quantitative methods, physicians make clinical decisions based on their patients’ symptoms. This has led to the development of a symptom-based classification system for diagnosis in clinical care and research (2). Within the FBD classification, there are certain disorders (e.g., irritable bowel syndrome [IBS], chronic functional abdominal pain, and painful constipation) where pain is the predominant complaint. With these disorders, physicians will also make diagnostic and treatment decisions based on the severity of the patient’s report of pain. Therefore, a patient with IBS having mild discomfort may receive little or no diagnostic evaluation, and a recommendation for diet modification. However, a patient with the same diagnosis who reports severe and disabling pain may receive extensive diagnostic testing and treatment (e.g., surgical or psychiatric consultation, invasive imaging studies such as endoscopic retrograde cholangiography, use of antidepressants, or experimental medications) (3, 4). The factors contributing to the patient’s report of pain in FBD, and to the physician’s diagnostic and therapeutic decisionmaking, are poorly understood. The purpose of this study was to assess the relative contributions of certain hypothesized clinical factors (psychosocial, behavioral, and physiological) to the severity of painful FBD. A better understanding of the determinants of severity in these conditions should lead to the development of more effective diagnostic and therapeutic strategies that are based on documented predictive factors.
AJG – April, 2000
Severity in Patients With Painful FBD
MATERIALS AND METHODS Study Sample We studied female patients, aged 18 – 65 yr, enrolling between 1996 and 1998 into a National Institutes of Health multicenter treatment trial for FBD at the University of North Carolina at Chapel Hill and the University of Toronto. The study subjects were required to have painful functional bowel symptoms (categorized as IBS, painful functional constipation, chronic functional abdominal pain, or unspecified painful functional bowel disorder) (1) for at least 3 months, and to have symptoms for at least 2 days of the week. We used a previously validated Functional Bowel Disorder Severity Index (FBDSI) (5) to define severity in the patient groups. Similar to the method used to develop the Crohn’s Disease Activity Index (6), the physicians’ global rating of severity for their patients was used as the dependent variable in a regression analysis. The clinical variables that best predicted this global rating in the final model were then used to determine the FBDSI. The FBDSI ⫽ [current pain by visual analog scale (0 –100)] ⫹ [diagnosis of chronic functional abdominal pain 0 if absent; 106 if present] ⫹ [number of physician visits over 6 months ⫻ 11]. For this study, patients with mild symptom severity (FBDSI ⬍37) were excluded from participation in the study. The remaining patients were stratified either into moderate (FBDSI ⫽ 37–110) or severe (FBDSI ⬎110) groups. Comparisons were made at study entry between moderate and severe patient groups with regard to the factors listed below. Measurement— Clinical and Psychosocial Factors Before and during the baseline evaluation period (and before entering treatment), the study subjects were given diary cards, self-report questionnaires, and underwent physiological investigation. For this study, the following measures were assessed: A) AVERAGED DAILY PAIN INTENSITY. Patients kept a daily diary for 2 wks before the study visit. Daily pain intensity was assessed using a 100-mm visual analog scale (0 ⫽ no pain, 100 ⫽ very severe), and average score for the 14 days was used in this analysis. B) PSYCHOSOCIAL QUESTIONNAIRES. At the time of the initial study visit, patients received the following standardized psychosocial questionnaires: 1. SCL-90 (7). The Global Symptom Index (GSI) was used as a state measure of psychological distress; 2. Beck Depression Inventory (BDI) (8). The BDI assessed symptoms of depression; 3. NEO Inventory (9). The Neuroticism scale of the NEO Inventory was used to assess personality dysfunction relating to resiliency in the face of stress; 4. Sickness Impact Profile (SIP) (10). The SIP is a generic health status instrument that focuses on daily function.
5.
6.
7.
8.
9.
975
We evaluated overall daily function (Overall Scale), in addition to the following subscales: Physical Function, Sleep and Rest, Home Management, Work, Recreation and Pastime, and Eating. Psychosocial function was not evaluated because it was too closely correlated with the other psychosocial measures. Social Support Questionnaire (SSQ-6) (11). The SSQ-6 measures social support using two subscales: Number of Supportive People, and Degree of Satisfaction with the support. Dysfunctional Attitudes Scale (DAS). The DAS is used to assess dysfunctional attitudes that may relate to symptom reporting, for example: “What other people think of me is very important.” Coping Strategies Questionnaire. (CSQ) (12). The CSQ is a coping measure frequently used for patients with chronic pain. We used specifically the Catastrophizing Scale of the CSQ and also two questions relating to subjects’ ability to control and decrease their symptoms. These items are predictive of poor health status in patients with functional GI disorders (13). IBS-QOL (14). The IBS-QOL is a well-validated condition-specific measure of quality of life for patients with IBS and other functional bowel disorders. Sexual/physical Abuse History (15). We used a confidential survey questionnaire from previous studies (15) to assess abuse history.
C) HEALTH CARE UTILIZATION/BEHAVIORAL MEASURES. We included the following: 1. Number of days in bed for GI symptoms over a 3-month period; 2. Number of overnight admissions (for all and just GI symptoms) over a 3-month period; 3. Number of total nightly stays (for all and just GI symptoms) over a 3-month period; 4. Number of total procedures (surgical, endoscopic) over the previous 3 months; 5. Number of M.D. visits (for all and just GI symptoms) over a 3-month period; 6. Number of times phoned GI physician over a 3-month period; 7. Number of hospitalizations in the last 2 yr; 8. Number of lifetime surgeries. D) STANDARDIZED RECTAL SENSATION THRESHOLDS. The Synectics Visceral Stimulator (MedTronic Synectics Medical, Shoreview, MN) was used to measure muscle tone and sensory thresholds by the barostat technique. Standardized assessment (16) of pressure and volume was obtained for: 1. 2. 3. 4.
Urge threshold (ascending method of limits); 1st pain report (ascending method of limits); Pain threshold (ascending method of limits; Pain threshold (tracking technique).
976
Drossman et al.
Statistical Analysis All data were entered into a data set using SAS Full Screen Editor, and analyzed using SAS (SAS, Cary, NC). Comparisons of demographic and clinical data were performed between study sites to assure comparability of the data for later analyses where data from these groups were combined. The UNC and Toronto data were used to compare moderate and severe patient groups for all except the physiology data, which was obtained from UNC only. We first performed GLM ANCOVA (adjusting for age, race, and education) for continuous variables, and 2 tests for categorical data to assess the relationship between the various factors with severity group status. After this, those factors that significantly discriminated between moderate and severe groups were entered as independent variables into a multiple logistic regression model, using severity group status (moderate vs severe) as the dependent variable. Because some psychosocial variables might be linearly associated, we performed collinearity testing to ascertain the appropriateness of the independent variables that were entered into the final logistic model. Although abdominal pain (VAS) score and number of MD visits/3 months discriminated moderate from severe groups, they were not included in the model, because these two variables are used to create the FBDSI. In addition, the SIP Overall Scale and the SIP Psychosocial subscale were not included because of collinearity effects, using condition indices ⬎15. The results from the multiple logistic regression are presented in terms of p values, adjusted odds ratios (OR), and 95% confidence intervals (CI).
RESULTS Study Population There were 211 female patients with FBD (108 Chapel Hill, 103 Toronto) who enrolled during the study period, of whom 84% were white (79% Chapel Hill, 89% Toronto; p ⬍ 0.04). The mean age was 39.3 ⫾ 12.1 SD, with 14.6 ⫾ 2.9 yr educational attainment. There were 144 patients (79 Chapel Hill, 65 Toronto) with moderate FBD and 67 patients (29 Chapel Hill, 38 Toronto) with severe FBD. The physicians identified 83.3% to have IBS, 3.3% painful functional constipation, 6.7% chronic functional abdominal pain, and 4.3% unspecified functional bowel disorder. There were no statistical differences between Chapel Hill and Toronto with regard to age, education, pain scores, and psychosocial and utilization/behavioral variables. Therefore, data from the two sites were combined in the later analyses. Comparison Between Moderate and Severe Groups DEMOGRAPHIC FACTORS. Patients with severe FBD had 1 yr less educational attainment (Moderate, 14.9 ⫾3.0 yr vs Severe, 13.9 ⫹ 2.5 yr; p ⬍ 0.01), and there were no
AJG – Vol. 95, No. 4, 2000
differences between groups with regard to race and age. Subsequent analyses were performed adjusting for age, race, and educational attainment. AVERAGED DAILY PAIN. Patients with severe FBD reported significantly greater pain (Severe, 51.0 ⫾ 2.5 vs Moderate, 35.1 ⫾ 1.7; p ⬍ 0.0001) on the 14-day averaged Visual Analog Scale score (0 ⫽ no pain, 100 ⫽ very severe pain). This pain measure was not used in further analyses, as pain is a defining variable for the FBDSI. PSYCHOSOCIAL SCORES. Table 1 compares the moderate and severe groups on the psychosocial scores, adjusting for age, race, and education. Patients with severe FBD reported significantly higher scores on the the Beck Depression Inventory (p ⬍ 0.005), the Overall Scale of the Sickness Impact Profile (p ⬍ 0.0001), and most subscales of the SIP (p ⬍ 0.01 to p ⬍ 0.0001). Patients with severe FBD also reported higher scores on the Catastrophizing Scale of the Coping Strategies Questionnaire (p ⬍ 0.0001), as well as less perceived ability to control (p ⬍ 0.05) or decrease (p ⬍ 0.01) symptoms. Finally, the severe FBD group showed significantly poorer quality of life on the Overall score on the IBS-QOL (p ⬍ 0.0003). There were no statistically significant differences with regard to SCL-90R GSI (trend at p ⬍ 0.07), neuroticism, social support, or dysfunctional attitudes. We also compared the frequency of physical and sexual abuse reported by questionnaire. There was a higher frequency of all types of abuse by the moderate group (57% vs 40%; p ⬍ 0.02). When evaluating the frequency of rape, the difference between groups was no longer statistically significant (22.2% vs 13%; p ⫽ 0.13). UTILIZATION AND BEHAVIORAL SCORES. Table 2 compares the moderate and severe groups with regard to the health care utilization and behavioral data, adjusting for age, race, and education. Patients with severe FBD reported more days in bed for GI symptoms (p ⬍ 0.0001), more overnight admissions for GI problems ⬎3 months (p ⬍ 0.03), more physician visits for GI and other symptoms ⬎3 months (p ⬍ 0.0001), more phone calls for GI symptoms ⬎3 months (p ⬍ 0.0001), and more hospitalizations in the previous 2 yr (0.0001). There was a trend toward more nightly stays in the hospital (p ⬍ 0.08) and more lifetime surgeries (p ⬍ 0.10). There was no difference between groups for the number of endoscopic and surgical procedure over the previous 3 months. RECTAL SENSATION THRESHOLDS. Table 3 compares the moderate and severe groups with regard to rectal thresholds to balloon distension using the barostat, adjusting for age, race, and education. There was a trend toward lower thresholds for the severe FBD group, but significance was achieved for only one of eight variables (volume at 1st pain report using the ascending method of limits; p ⬍ 0.05).
AJG – April, 2000
Severity in Patients With Painful FBD
977
Table 1. Comparison Between Moderate and Severe FBD: Psychosocial Tests Psychosocial Scale
Moderate FBD Mean (S.E.)
Severe FBD Mean (S.E.)
p⬍
SCL-90 (GSI) Beck Depression Index NEO (Neuroticism) SIP Overall Score SIP Physical SIP Sleep/Rest SIP Home Management SIP Work SIP Recreation SIP Eating Social Support Questionnaire SSQ # People SSQ Satisfaction Dysfunctional Attitudes Perfectionism Approval by others Coping Strategies Questionnaire Catastrophizing Control of symptoms Decrease of symptoms IBS-QOL Overall
57.6 (3.5) 9.3 (0.7) 89.2 (2.0) 5.5 (0.5) 2.2 (0.3) 10.2 (1.0) 5.5 (0.9) 9.8 (1.9) 13.1 (1.3) 2.0 (0.4)
69.3 (5.2) 12.5 (1.0) 91.4 (3.0) 9.3 (0.8) 5.4 (0.5) 13.6 (1.6) 10.8 (1.3) 18.5 (2.9) 18.7 (1.9) 5.1 (0.5)
0.07 0.005 0.54 0.0001 0.0001 0.07 0.001 0.01 0.001 0.0001
21.1 (1.0) 30.6 (0.5)
19.1 (1.5) 30.4 (0.7)
0.28 0.84
83.3 (1.3) 44.5 (0.8)
86.2 (1.9) 44.0 (1.2)
0.20 0.75
8.2 (0.6) 2.6 (0.1) 2.4 (0.1) 69.3 (1.6)
12.9 (0.8) 2.2 (0.2) 1.9 (0.2) 58.8 (2.4)
0.0001 0.05 0.01 0.0003
FBD ⫽ functional bowel disorders; SIP ⫽ Sickness Impact Profile. For explanation of psychosocial scales used, see Materials and Methods section.
Multiple Logistic Regression
DISCUSSION
To determine the relative importance of the factors that best predict severity, a logistic regression model was fitted with Severity (Moderate vs Severe) as the dependent variable, and the significant variables from the previous ANCOVAs (Tables 1–3) were considered in a stepwise fashion as independent predictor variables. Table 4 shows the results from the final logistic regression. Of the 19 variables entered into the model, only four—Physical Dysfunction (SIP) (p ⬍ 0.01), Eating Dysfunction (SIP) (p ⬍ 0.027), days in bed for GI symptoms (p ⬍ 0.005), and number of times phoned M.D. for GI symptoms (p ⬍ 0.009) significantly predicted severity. We also ran a separate regression analysis including the visceral sensitivity data as predictor variables for the UNC data set and the visceral sensitivity variables were not significant predictors (data not reported).
For the irritable bowel syndrome (IBS) and other functional bowel disorders (FBD), diagnosis and treatment usually depend on the predominant symptom pattern. Thus, patients with predominant diarrhea will have different diagnostic studies and treatments than patients with predominant constipation or abdominal bloating (4). This has led to the development and use of a symptom-based diagnostic classification system, the Rome criteria. (2) However, when pain is a dominant feature of the condition, physicians also make these decisions based on another clinical dimension, the severity of the pain reported (3, 4, 17). Reports of pain that is mild in intensity may lead to little or no diagnostic evaluation, and treatment may include simple reassurance or a recommendation for diet and lifestyle change. But reports of severe pain may lead to extensive diagnostic eval-
Table 2. Comparison Between Moderate and Severe FBD: Health Care Utilization and Behaviors Health Care/Behaviors
Moderate FBD Mean (S.E.)
Severe FBD Mean (S.E.)
p⬍
Days in bed for GI/3 months Overnight admissions/3 months Overnight admissions GI/3 months Total nightly stays/3 months Total nightly stays GI/3 months Total procedures Total MD visits/3 months Total MD visits GI/3 months Times phoned GI MD/3 months Hospitalizations/2 yr Lifetime surgeries
2.2 (0.9) 0.02 (0.02) 0.002 (0.01) 0.02 (0.2) 0.01 (0.08) 0.08 (0.03) 2.1 (0.40) 1.1 (0.3) 0.4 (0.2) 0.2 (0.09) 2.4 (0.2)
11.8 (1.4) 0.06 (0.03) 0.06 (0.02) 0.5 (0.2) 0.03 (0.1) 0.05 (0.04) 6.9 (0.5) 4.8 (0.4) 1.6 (0.2) 0.9 (0.1) 3.1 (0.4)
0.0001 0.35 0.03 0.08 0.08 0.62 0.0001 0.0001 0.0001 0.0001 0.10
FBD ⫽ functional bowel disorders.
978
Drossman et al.
AJG – Vol. 95, No. 4, 2000
Table 3. Comparison Between Moderate and Severe FBD: Rectal Sensitivity Rectal Sensitivity Parameter Pain threshold Pressure Volume First pain report Pressure Volume Urge threshold Pressure Volume Pain threshold (tracking) Pressure Volume
Moderate FBD Mean (S.E.)
Severe FBD Mean (S.E.)
p⬍
30.0 (1.3) 164.6 (11.2)
27.6 (2.2) 125.9 (18.6)
0.36 0.08
32.6 (3.1) 170.7 (11.1)
23.8 (5.2) 127.9 (18.5)
0.16 0.05
27.0 (1.9) 151.9 (10.7)
22.0 (3.1) 119.3 (17.7)
0.17 0.12
29.9 (4.0) 195.6 (12.4)
22.9 (6.6) 149.1 (20.5)
0.38 0.06
FBD ⫽ functional bowel disorders.
uation (e.g., colonoscopy, computed tomography [CT] scan or endoscopic retrograde cholangiopancreatography [ERCP] and biliary manometry, or surgical or psychiatric consultation), and treatments (e.g., prescription of antidepressants, or employment of experimental medications and alternative treatments). Recent studies support an important role for psychosocial factors in influencing pain severity among patients with functional GI disorders. These factors may amplify the pain experience and/or increase illness-related behaviors that communicate the pain to others (4, 18). Although intestinal dysmotility contributes to many functional GI symptoms (e.g., diarrhea, constipation, vomiting), there is little evidence to show that it contributes to chronic GI pain. However, there is an association between visceral hypersensitivity and pain reports (19), which may be mediated at the level of the afferent receptor, the spinal cord, or higher centers, and current opinion favors a strong contribution from the central nervous system (CNS) (20, 21). In fact, brain-imaging studies implicate a regulatory role for the neocortex (e.g., anterior cingulate gyrus, prefrontal cortex) in the affective experience of visceral pain (22, 23). Finally, psychosocial disturbance may influence the development of functional GI symptoms after an acute enteric infection (24). This may occur through stress-mediated effects both on motility and also on inflammation via the HPA immune/ cytokine axis (25). Inflammation (possibly resulting from CNS activation of peripheral cytokines) can lead to intestiTable 4. Stepwise Logistic Regression: Final Model Predictor Variable*
Odds Ratio
95% CI
p⬍
SIP Physical Subscale SIP Eating Days in bed for GI/3 months # Phone calls to MD/3 months
0.87 0.90 0.92 0.67
(0.78–0.97) (0.82–0.99) (0.86–0.97) (0.48–0.88)
0.013 0.027 0.005 0.009
*The following variables were also entered into the stepwise logistic regression, but were eliminated from the final model: Age, race, education, Beck Depression Inventory, SCL-90 Global Severity, Catastrophizing Scale of CSQ, perceived ability to decrease or control symptoms (CSQ), IBS-QOL Overall Score, SIP Sleep and Rest, SIP Home Management, SIP Work, SIP Recreation, Days in bed for GI symptoms (3 months), # Hospitalization (2 yr), # Lifetime surgeries.
nal dysmotility and visceral sensitization, even after the acute infection resolves (26). The aim of the current study was to determine the factors that contribute to severity among patients with painful functional bowel disorders (IBS, painful functional constipation, and chronic functional abdominal pain). Our hypothesis was that psychosocial, and behavioral, more than physiological (visceral hypersensitivity) factors would predict severity, as determined by the Functional Bowel Disorder Severity Index (FBDSI) (5). As predicted, we found that most of the psychosocial and behavioral variables significantly discriminated these two patients groups, and visceral hypersensitivity had little or no effect. Furthermore, the logistic regression indicated that severity was best associated with patient behaviors rather than with psychosocial factors: their daily physical functional status and eating capability, the number of days spent in bed, and the number of phone calls to physicians because of GI illness. Why might these behaviors be such strong predictors of severity? We believe that the physician’s assessment of severity in FBD is influenced primarily by how the patient behaves in response to the pain (27). As previously shown with the development of the FBDSI, the physician’s global rating of pain was best determined by the severity and chronicity of the patient’s report of pain, and the number of physician visits. Furthermore, in the current study, the severe group was characterized by additional patient behaviors: restricting or modifying the diet because of pain (SIPEating subscale), frequently phoning the physician (number of phone calls/3 months), cutting back on daily physical activities (SIP-Physical subscale), and taking to bed (number of days in bed/3 months for GI problems). Although the physician perceives these behaviors as reflective of severe physical symptoms, a dilemma ensues, because, in many cases, diagnostic and therapeutic efforts directed toward the symptoms fail. This dilemma may result from how physicians understand and respond to patients with chronic pain behaviors. When applying the traditional biomedical model, physicians will search for a specific medical source for the pain and will employ medical treatments, expecting symptom relief (28). However, this expectation is not usually met: expensive diagnostic evaluations do not disclose other conditions (27), and treatments directed at the gut (e.g., antispasmodics, antidiarrheal drugs, etc.) are often unsuccessful (17). Furthermore, the underlying psychosocial factors contributing to the patient’s pain experience and behavior, as identified in this study, are usually ignored, minimized, or relegated to being secondary phenomena. As shown in one study of medical outpatients with IBS (29), physicians could more accurately identify the details of the patients’ somatic complaints than they could patients’ symptom-related cognitions, behavior, and anxieties, all of which are important components of the experience of pain. The result is that many patients with painful functional GI symptoms become dissatisfied with the care they receive,
AJG – April, 2000
and may seek the advice of other physicians or obtain alternative treatments (30). In turn, the physicians remain perplexed by the patients’ continued complaints despite no evidence for “disease.” They are frustrated in their unsuccessful efforts, may assume that the patient cannot be helped, and may develop rejecting or stigmatizing attitudes and behaviors. Ultimately this leads to a deterioration in the physician-patient relationship, and the perpetuation of a vicious cycle of refractoriness and help-seeking. For these reasons, we conclude that the biomedical model is neither satisfactory nor sufficient for patients with FBD. We would recommend a biopsychosocial approach in which the symptoms are understood in terms of the biological and psychosocial determinants (28). The findings of this study have important implications with regard to the application of this model in the care of patients with painful FBD. First, efforts can be made to reduce the pain (via central analgesia) and possibly psychiatric comorbidity (e.g., depression, panic) through the use of antidepressants. Second, efforts can also be made to modify pain-related behaviors, and increase cognitive coping strategies through referral to a mental health professional skilled in psychological techniques (e.g., interpersonal psychotherapy or cognitive therapy), which have been shown effective. Preliminary studies suggest that both antidepressants and psychological treatments improve the clinical outcome of patients with functional GI disorders, but further well-designed studies of large patient samples are needed. Finally, there is a compelling need to educate physicians on how to elicit relevant psychosocial data and integrate this information into a comprehensive, biopsychosocial plan of care. The physician can achieve this by (31) demonstrating a genuine interest in the patient’s beliefs and expectations relating to the condition; using patient-centered communication strategies (32); eliciting the patient’s concerns and offering appropriate reassurance; understanding and seeking to modify any somatic misattributions (e.g., that the patient has cancer); making educated decisions about the use of psychopharmacologic agents; and, when appropriate, arranging mental health referral concurrent with the medical care. It was recently shown that diagnostic investigations among patients with IBS did not improve patient satisfaction. However, when the physicians correctly identified and addressed the patients’ psychological attributions related to their IBS, their satisfaction with the care increased, and level of anxiety, and maladaptive cognitions diminished (33). Recent recommendations have been made by the American Gastroenterological Association to provide supervised time for clinical and psychosocial skill training for fellows seeing patients with functional GI disorders (34). There are three features relating to the study design or analysis that deserve further comment. First, the study involved patients with moderate to severe FBD seen mostly in referral (medical center) practices. Therefore, the conclusions drawn from these data may not apply to the majority of patients seen in primary care practice or to patients with
Severity in Patients With Painful FBD
979
milder or nonpainful symptoms. Patients with mild FBD were excluded by the protocol to understand this more difficult-to-treat group of patients. Second, we may have overlooked other important predictor variables that might contribute to severity. These might involve additional psychosocial (e.g., psychiatric or personality trait) or medical (e.g., stool frequency, vomiting, etc.) variables. Nevertheless, the available data compel us to conclude that psychosocial and behavioral factors are strong determinants of severity. Finally, in comparing differences between moderate and severe groups we did not adjust for multiple comparisons. Our preliminary (ANCOVA) analyses were descriptive in nature, and were also used to identify meaningful variables for the logistic regression.
CONCLUSIONS We conclude that patient illness behaviors are the best predictors of severity in FBD. The use of psychopharmacological agents (e.g., antidepressants) and psychological treatments may help reduce patient distress through treatment of psychological comorbidity and improvement in coping styles. In addition, training to help medical physicians identify and respond to psychosocial and behavioral features is likely to improve patients’ satisfaction with the care and the clinical outcome.
ACKNOWLEDGMENTS This study was supported by NIH Grants #RO1DK49334 and KO5MH00133, UNC Center for Functional GI and Motility Disorders, the UNC Center for Gastrointestinal Biology and Disease, and the UNC General Clinical Research Center. We thank Udi Ringel, M.D., for his review of this article. Reprint requests and correspondence: Douglas A. Drossman, M.D., Division of Digestive Diseases, Department of Medicine, University of North Carolina, 726 Burnett-Womack Building, CB #7080, Chapel Hill, NC 27599-7080. Received Apr. 29, 1999; accepted Sep. 7, 1999.
REFERENCES 1. Thompson WG. C. Functional bowel disorders and D. Functional abdominal pain. In: Drossman DA, Richter JE, Talley NJ, et al., eds. Functional gastrointestinal disorders: Diagnosis, pathophysiology and treatment, 1st ed. McLean, VA: Degnon Associates, 1994:115–73. 2. Drossman DA, Richter JE, Talley NJ, et al. Functional gastrointestinal disorders: Diagnosis, pathophysiology and teatment, 1st ed. McLean, VA: Degnon Associates, 1994:1–370. 3. Drossman DA, Thompson WG. The irritable bowel syndrome: Review and a graduated, multicomponent treatment approach. Ann Intern Med 1992;116:1009 –16. 4. Drossman DA, Whitehead WE, Camilleri M. Irritable bowel syndrome: A technical review for practice guideline development. Gastroenterology 1997;112:2120 –37.
980
Drossman et al.
5. Drossman DA, Li Z, Toner BB, et al. Functional bowel disorders. A multicenter comparison of health status, and development of illness severity index. Dig Dis Sci 1995;40:986 –95. 6. Best WR, Becktel JM, Singleton JW, et al. Development of a Crohn’s disease activity index. National Cooperative Crohn’s disease study. Gastroenterology 1976;70:439 – 44. 7. Derogatis LR. SCL-90-R: Administration, scoring, and procedures manual II —For the r(evised) version, 1st ed. Towson, MD: Clinical Psychometric Research, 1983. 8. Beck AT, Ward CH, Mendelson M, et al. An inventory for measuring depression. Arch Gen Psychiatry 1961;4:561–71. 9. Costa PT Jr, McCrae RR. The NEO personality inventory manual. Odessa, FL: Psychological Assessment Resources, 1985. 10. Bergner M, Bobbitt RA, Carter WB. The sickness impact profile: Development and final revision of a health status measure. Med Care 1981;19:787– 805. 11. Sarason IG, Sarason BR, Shearin EN, et al. A brief measure of social support: Practical and theoretical implications. J Soc Personal Relationships 1987;4:497–510. 12. Rosenstiel AK, Keefe FJ. The use of coping strategies in chronic low back pain patients: Relationship to patient characteristics and current adjustment. Pain 1983;17:33– 40. 13. Drossman DA, Li Z, Leserman J, et al. Effects of coping on health outcome among female patients with gastrointestinal disorders. Psychosom Med 2000 (in press). 14. Patrick DL, Drossman DA, Frederick IO, et al. Quality of life in persons with irritable bowel syndrome: Development of a new measure. Dig Dis Sci 1998;43:400 –11. 15. Drossman DA, Leserman J, Nachman G, et al. Sexual and physical abuse in women with functional or organic gastrointestinal disorders. Ann Intern Med 1990;113:828 –33. 16. Whitehead WE, Delvaux M, Working Team. Standardization of procedures for testing smooth muscle tone and sensory thresholds in the gastrointestinal tract. Dig Dis Sci 1994;42: 223– 41. 17. Drossman DA. Diagnosing and treating patients with refractory functional gastrointestinal disorders. Ann Intern Med 1995;123:688 –97. 18. Drossman DA, Creed FH, Olden KW, et al. Psychosocial aspects of the functional gastrointestinal disorders. Gut 1999; 45:II25–II30. 19. Mertz H, Naliboff B, Munakata J, et al. Altered rectal perception is a biological marker of patients with irritable bowel syndrome. Gastroenterology 1995;109:40 –52.
AJG – Vol. 95, No. 4, 2000
20. Mayer EA, Gebhart GF. Basic and clinical aspects of visceral hyperalgesia. Gastroenterology 1994;107:271–93. 21. Whitehead WE, Palsson OS. Is rectal pain sensitivity a biological marker for irritable bowel sydrome: Psychological influences on pain perception. Gastroenterology 1998;115:1– 10. 22. Silverman DHS, Munakata JA, Ennes H, et al. Regional cerebral activity in normal and pathologic perception of visceral pain. Gastroenterology 1997;112:64 –72. 23. Aziz Q, Thompson DG. Brain-gut axis in health and disease. Gastroenterology 1998;114:559 –78. 24. Gwee KA, Leong YL, Graham C, et al. The role of psychological and biological factors in post-infective gut dysfunction. Gut 1999;44:400 – 6. 25. Drossman DA. Mind over matter in the postinfective irritable bowel. Gut 1999;44:306 –7. 26. Collins SM. The immunomodulation of enteric neuromuscular function: Implications for motility and inflammatory disorders. Gastroenterology 1996;111:1683–99. 27. Drossman DA. Chronic functional abdominal pain. Am J Gastroenterol 1996;91:2270 – 81. 28. Drossman DA. Presidential address: Gastrointestinal illness and biopsychosocial model. Psychosom Med 1998;60:258 – 67. 29. Van Dulmen AM, Fennis JFM, Mokkink HGA, et al. Doctors’ perception of patients’ cognitions and complaints in irritable bowel syndrome at an out-patient clinic. J Psychosom Res 1994;38:581–90. 30. Sutherland LR, Verhoef MJ. Why do patients seek a second opinion or alternative medicine? J Clin Gastroenterol 1994;3: 194 –7. 31. Drossman DA. Psychosocial factors in the care of patients with gastrointestinal disorders. In: Yamada T, ed. Textbook of gastroenterology, 3rd ed. Philadelphia: Lippincott-Raven, 1999:638 –59. 32. Drossman DA. Psychosocial sound bites. Exercises in the patient-doctor relationship. Am J Gastroenterol 1997;92: 1418 –23. 33. Van Dulmen AM, Fennis JFM, Mokkink HGA, et al. Doctordependent changes in complaint-related cognitions and anxiety during medical consultations in functional abdominal complaints. Psychol Med 1995;25:1011– 8. 34. Ouyang A, Camilleri M, Drossman DA, et al. Task force on training in motility, diverticular disease, and functional illness. Gastroenterology 1996;110:1274 – 6.