SEARCH

SEARCH BY CITATION

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

Aim To provide estimates of actual costs to deliver health care to patients with functional bowel disorders, and to assess the cost impact of symptom severity, recency of onset, and satisfaction with treatment.

Methods We enrolled 558 irritable bowel (IBS), 203 constipation, 243 diarrhoea and 348 abdominal pain patients from primary care and gastroenterology clinics at a health maintenance organization within weeks of a visit. Costs were extracted from administrative claims. Symptom severity, satisfaction with treatment and out-of-pocket expenses were assessed by questionnaires.

Results Average age was 52 years, 27% were males, and 59% participated. Eighty percent were seen in primary care clinics. Mean annual direct health care costs were $5049 for IBS, $6140 for diarrhoea, $7522 for constipation and $7646 for abdominal pain. Annual out-of-pocket expenses averaged $406 for treatment of IBS symptoms, $294 for diarrhoea, $390 for constipation and $304 for abdominal pain. Lower gastrointestinal costs comprised 9% of total costs for IBS, 9% for diarrhoea, 6.5% for constipation and 9% for abdominal pain. In-patient care accounted for 17.5% of total costs (15.2% IBS).

Conclusion Costs were affected by disease severity (increased), recent exacerbation of bowel symptoms (increased), and whether the patient was consulting for the first time (decreased).


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

Irritable bowel syndrome (IBS) is a highly prevalent disorder affecting an estimated 10–15 percent of the population.1 In addition, IBS patients are reported to utilize health care resources disproportionately to the seriousness of their symptoms.2 These factors of high prevalence and high costs per patient combine to make IBS a significant source of health care costs and a potential target for reducing health care expenditures. The potential for cost savings has motivated several research groups to investigate health care costs in IBS. However, total cost and component cost estimates vary widely among studies, both within the United States and between countries. Reasons for this variability in estimates include differences in ascertainment criteria (specialty care or primary care patients or community samples) and methods for calculating costs.3

In this study, we took advantage of a large health maintenance organization [Group Health Cooperative of Puget Sound (GHC)] with a well-developed automated medical records system and outcomes research infrastructure that enabled us to capture health care costs from a representative sample of IBS and other functional gastrointestinal (GI) patients in both primary and gastroenterology clinics. Through a prospective study design, we combined a 24-month administrative claims review with a 6-month chart review. We sent postal questionnaires to patients in order to capture health care costs paid out of pocket that are missed in many cost of care studies. We also gathered questionnaire data that allowed us to relate health care costs to IBS symptom severity, success of treatment, and whether the patient was consulting about their bowel symptoms for the first time.

The aims of this study were to estimate total and component costs for four groups of patients – IBS, functional diarrhoea, functional constipation and functional abdominal pain – in a setting that is representative of staff model health maintenance organizations (HMOs) in the United States. Specifically, we aimed to compare: (i) total direct health care costs of the four patient groups, (ii) costs per group for each of the four 6-month periods covered by our study, (iii) cost components for each group, (iv) costs of first-time consulters compared to returning patients, (v) costs according to IBS symptom severity and (vi) costs of IBS patients who expressed satisfactory relief compared to those who did not report satisfactory relief from their symptoms.

Methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

Setting

The Managing Abdominal Pain Study4 was carried out at GHC, a large staff model HMO in Seattle with more than 525 000 members. The demographics of this patient population are similar to the demographics of the Seattle area5 except for less representation of the highest and lowest income groups. Compared to the rest of the United States, GHC and the Seattle area also have a higher proportion of Asians and a lower proportion of African-Americans and Hispanics, but the ethnicity of GHC membership is otherwise similar to the United States population. Physicians who cared for the patients participating in this survey included 353 primary care physicians and 16 gastroenterology specialists. Patients were enrolled in this survey between December 4, 2001 and September 12, 2002.

Inclusion criteria

To identify patients with IBS and related functional bowel symptoms, all patient encounter forms submitted by physicians in primary care and gastroenterology clinics were prospectively screened for IBS (ICD-9CM code 564.1), abdominal pain (789.X), constipation (564.0), or diarrhoea (787.91). This process identified patients soon after they had consulted for diagnosis and treatment of a GI complaint; this was their index visit.

To be eligible for the study, participants had to be age 18–75, enrolled in GHC for at least one year, and have no visit record with a diagnosis of ulcerative colitis or Crohn’s disease. Sampling was stratified with the goal of enrolling 500 IBS, 300 abdominal pain, 200 constipation and 200 diarrhoea patients from primary care clinics as well as 300 IBS, 200 abdominal pain, 100 constipation and 100 diarrhoea patients from gastroenterology clinics.

Design

This study was carried out as part of a larger investigation of standard medical care for IBS4 Subjects who met inclusion criteria were mailed an invitation to participate in the study, an informed consent statement, and the first set of questionnaires, usually within 2 weeks of their index visit. In the invitation to participate, subjects were told they could telephone or write to the survey team at GHC if they did not want to participate (refused). They were also informed that if they neither returned the questionnaires nor refused, they would be telephoned after approximately 2 weeks to remind them to complete the questionnaires. At least six attempts were made to contact each participant if they were not contacted before this point. The invitation to participate also informed subjects that they would be re-contacted in 6 months to collect follow-up information on their medical care and that we were requesting permission to review their medical records. Only subjects who returned the initial set of questionnaires were mailed follow-up questionnaires. Subjects were offered a $10 incentive for completing each of the surveys. This study was reviewed and approved by the institutional review boards of GHC, the University of Washington and the University of North Carolina at Chapel Hill.

Administrative claims

Administrative claims at the HMO were reviewed for various categories of direct costs of care for four time periods: Period 1 (12 months to 6 months prior to the index visit), Period 2 (6 months prior to the index up to the index visit), Period 3 (from the index visit to 6 months later) and Period 4 (from 6 months to 12 months after the index period). Major cost components are described in Table 1. For each cost category, lower GI costs were recorded separately from non-GI costs; however, total costs include both lower GI and all other costs for that category.

Table 1.   Major all-cause cost components
Cost variableDescription
  1. * Data for these expenditures were gathered through questionnaires described below.

In-patient
 In-patient costsAll in-patient costs, except minor costs for mental health or alcohol/drug abuse treatment (not including long-term care or alcohol/drug abuse treatment)
Out-patient costs
 Emergency Emergency care services
 LabLaboratory tests, including endoscopies and chemistries
 Mental health All out-patient mental health services
 Primary care office visitAll costs for a visit to a primary care clinic
 Radiology Radiology tests
 PharmacyAll out-patient pharmacy costs captured by the HMO’s administrative claims
 Specialist care office visitAll costs for a visit with a gastroenterologist or other medical specialist
 OtherAny remaining costs not captured in specific categories, including community health, hospital out-patient, contact lens services, drug or alcohol abuse treatment, and long-term care
Out-of-pocket*
 Non-prescription medicationsAcid reducers, laxatives, antidiarrheals, stool softeners, gas relief medications, pain medications, antispasmodics, other medications, fibre, bran, special diet, exercise program
 Alternative treatmentsGinger root or tea, fennel seed, senna tea, psychotherapy, megavitamins, homeopathic methods, special diet, hypnotherapy, massage therapy, biofeedback, acupuncture, yoga, aromatherapy

Chart review in Period 3

For patients who consented to have their medical charts reviewed (93% of those who completed enrolment surveys), trained chart abstractors reviewed the charts for the 6-month period between the index visit and the follow-up survey (Period 3). Patients with a functional bowel diagnosis who subsequently received a diagnosis of GI malignancy, inflammatory bowel disease (IBD) or malabsorption (including celiac disease) were excluded from further analyses. Excluded on this basis were 15 IBS patients, 29 diarrhoea patients, seven constipation patients and 21 abdominal pain patients. The reliability of chart reviewer judgments was tested by having two independent chart reviewers separately rate 101 charts, and their agreement was 91% for IBS diagnosis.

Questionnaires

The initial packet of questionnaires included the following: (a) Rome questionnaire to determine whether the subject met Rome II criteria;6 (b) the Irritable Bowel Severity Scale (IBSS),7 which is a validated questionnaire for determining the overall severity of IBS symptoms; (c) the Irritable Bowel Syndrome Quality of Life Scale (IBS-QOL);8 (d) a newly developed questionnaire which asked patients which treatments their doctors recommended and requested them to rate their confidence that these treatments would help them; and (e) the Brief Symptom Inventory-18 (BSI),9 which assesses psychological symptoms.

Questionnaires given at follow-up (6 months after the index visit) included: (a) a binary response question assessing whether the patient had obtained satisfactory relief of bowel symptoms; (b) a 7-step ordinal scale rating of improvement in GI symptoms;10 and (c) questions on how well the patient adhered to specific treatment recommendations. The following scales were re-administered to assess changes in health status since enrolment: (a) the IBSS to assess symptom severity, (b) the IBS-QOL to assess IBS-specific quality of life and (c) the BSI to assess psychological symptoms.

Out-of-pocket expenditures

The questionnaires completed by participants at 6 months included a questionnaire that asked about out-of-pocket expenditures for treatment of bowel symptoms within the last 3 months, in $10 increments from $10 to $100 and over. Non-prescription and alternative therapies included in the questionnaire are listed in Table 1. Costs reported in the previous 3 months were multiplied by 2 to provide an estimate of out-of-pocket costs for Period 3. When estimating annual out-of-pocket expenditures, the costs reported in the questionnaire were multiplied by 4. We report out-of-pocket expenses separately from costs incurred through GHC (administrative claims data), because the payers are different – a matter of importance to health economists.

Data analysis

This is a descriptive study, where the data are presented as mean costs. Where comparisons are made between groups, we used paired samples t-test or general linear model (GLM), with the alpha level set at 0.05. All reported means are the arithmetic means, however for the purposes of statistical testing, we transformed each cost category to log10 to normalize the distribution of values. All analyses were conducted using SPSS Version 15.0 for Windows.

More detailed data for all analyses are provided in the on-line version of the manuscript (http://www.blackwell-synergy.com). These supplemental tables provide means, 95% confidence intervals, standard deviations, ranges, skew and kurtosis.

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

Patient characteristics

Of the 3024 initial surveys mailed, 20 actively refused and 1218 passively refused by not returning the survey despite reminder phone calls, and 1770 completed the questionnaires (59%). As already noted, we excluded from our analysis patients identified as having cancer, IBD or malabsorption (relatively rare but costly diseases) and a small group of 72 healthy controls. The remaining 1352 patients were diagnosed by a physician as having IBS (41%), diarrhoea (18%), constipation (15%), or abdominal pain (26%). The demographics of the sample are presented in Table 2.

Table 2.   Demographics of the sample
GroupProviderClinical diagnosisTotal
PCGIIBSAbdominal painConstipationDiarrhoea
  1. PC, primary care; GI, gastrointestinal; IBS, irritable bowel syndrome.

  2. * Statistically significant difference between patients seen in GI clinics vs. primary care (P < 0.05).

  3. † Statistically significant difference between patients with different clinical diagnoses (P < 0.05). Groups that differed from each other are identified by †.

Number1081271*5583482032431352
Gender (% male)27.323.622.8027.6031.5029.6026.60
Age (mean years)51.8054.75*50.92†53.2054.13†53.1552.39
Ethnicity (% hispanic)4.023.863.562.145.566.223.99
Race:
 % Caucasian89.2485.3489.4587.2885.5090.3088.45
 % African-American3.213.762.185.035.501.693.32
 % Asian2.936.02*2.184.735.503.383.55
Married or cohabitating (%)71.7674.9172.8873.1269.9572.3172.40
College graduate self (%)41.9449.45*45.9340.2938.9246.0943.45
College graduate partner (%)41.5149.77*48.51†38.32†38.2242.2943.22

In our sample, 20% of the patients were seen in specialty GI clinics and 80% in primary care clinics. Compared to primary care patients, gastroenterology patients were older [F(1,1347) = 9.217, P < 0.05] and included proportionately more Asian patients [F(1,1323) = 5.942, P < 0.05]. Gastroenterology patients were also more likely to be college graduates [F(1,1342) = 4.975, P < 0.05] and to have a partner who is a college graduate [F(1,1067) = 4.892, P < 0.05].

Comparing patients with different clinical diagnoses, IBS patients were significantly younger than constipated patients [F(3,1345) = 3.566, P < 0.05]. IBS compared to abdominal pain patients also had a higher proportion of partners who were college graduates [F(3,1065) = 3.129, P < 0.05].

Effects of sex, age, race and education on health care costs

Separate analyses were conducted on the demographic variables gender, age (young = 39 years and below, middle aged = 40–59 years, old = 60 years and above), race (Caucasian vs. all other), and educational achievement (college degree vs. no college degree). Education was a surrogate marker for socioeconomic status. For all diagnostic groups combined, total health care costs were unrelated to sex or race, but increased significantly (and linearly) with advancing age [F(3,1251) = 23.666, P < 0.05]. Total health care costs were also significantly lower for college graduates compared to those with less education [F(1,1245) = 21.407, P < 0.05]. This pattern was repeated among IBS patients, with significant differences for age [F(3,512) = 10.382, P < 0.05] and education [F(1,509 = 15.831, P < 0.05]. Demographic differences in costs were not tested for other diagnostic groups separately due to small sample size, but descriptive statistics for all groups are given in Table S1 of the on-line version of the paper (http://www.blackwell-synergy.com).

For lower GI costs only, when all diagnostic groups were combined, there were no significant differences related to sex or race. Lower GI costs were significantly lower for college graduates [F(1,1245) = 9.914], similar to total costs. However, when age effects on lower GI costs were examined, middle age patients incurred significantly higher costs than either young or old patients who were not different from each other [F(3,1251) = 3.323, P < 0.05]. For IBS patients only, there were no statistically significant differences related to these demographic variables. See Table S2 for detailed data (http://www.blackwell-synergy.com).

Finally, out-of-pocket costs were significantly higher for women compared to men [F(1,924) = 11.746, P < 0.05], and for college graduates compared to non-graduates [F(1,920) = 4.440, P < 0.05]. However, out-of-pocket expenses were not related to age or race. For IBS patients only, costs again were higher for female patients compared to male patients [F(1,392) = 4.099, P < 0.05] and for college graduates compared to non-graduates [F(1,389) = 5.663, P < 0.05], but there were no differences related to age or race. See Table S3 for detailed data (http://www.blackwell-synergy.com).

Total annual costs by group

Mean total annual costs for health care provided through Group Health Cooperative (claims data) were $5049 (CI $4441–$5667) for IBS, $6140 (CI $5070–$7151) for diarrhoea, $7522 (CI $5689–$9146) for constipation and $7646 (CI $6458–$8736) for abdominal pain patients (Figure 1). Total costs were significantly different between diagnostic groups [F(3,1251 = 13.657, P < 0.05]. Post hoc comparisons showed that total costs were significantly lower for IBS and diarrhoea compared to abdominal pain, and IBS costs were significantly lower compared to constipation.

image

Figure 1.  Total annual costs (mean). bsl00023Out-of-pocket bsl00001Claims data. Asterisk indicates abdominal pain group was significantly different from the irritable bowel syndrome (IBS) and diarrhoea groups (P < 0.05).

Download figure to PowerPoint

When patients were classified based on Rome II criteria rather than clinical diagnosis, there were no significant cost differences between diagnostic groups, but trends were similar: total costs tended to be higher for functional constipation patients compared to IBS and functional diarrhoea patients (See Table S5 in the on-line version of the study).

Annual estimates for total out-of-pocket expenses are also shown in Figure 1. These costs added $406 (CI $359–$454) for IBS, $294 (CI $242–$347) for diarrhoea, $390 (CI $320–$461) for constipation and $303 (CI $259–$347) for abdominal pain. Out-of-pocket expenditures averaged $357 (5.7% of total health care expenditures).

Total costs by period

Costs were highest in Period 3 for all diagnostic groups (see Figure 2). For each diagnostic group, costs in Period 3 were significantly higher relative to Periods 2 and 4.

image

Figure 2.  Average total costs by 6-month period, by group (mean). □Period 1 bsl00023Period 2 bsl00001Period 3 bsl00022Period 4. Period 3 was significantly greater than periods 2 and 4 for all groups (P < 0.05). IBS, irritable bowel syndrome.

Download figure to PowerPoint

Lower GI costs

Lower GI costs comprised a relatively small proportion of total annual GHC costs for each group: 9% of total annual costs for IBS ($456), 7% for diarrhoea ($416), 6.5% for constipation ($487) and 9% for abdominal pain ($687) (see Figure 3). Lower GI costs for abdominal pain patients were significantly higher compared to all other diagnostic groups [F(3,1251) = 18.082, < 0.05]. Lower GI costs were the highest in Period 3 for all diagnostic groups, averaging 11.6% of mean total costs.

image

Figure 3.  Total lower GI annual costs, by group(mean). bsl00023Out-of-pocket bsl00001Claims data. Total lower GI costs for the abdominal pain group were significantly higher than irritable bowel syndrome (IBS) and diarrhoea by paired comparisons (P < 0.05). IBS, diarrhoea and constipation were not different from each other. GI, gastrointestinal.

Download figure to PowerPoint

All out-of-pocket expenditures for non-prescription drugs and alternative medicines were ascribed to lower GI costs because the questionnaire asked subjects about the use and cost of treatments for bowel symptoms. When these out-of-pocket expenditures were added to lower GI costs incurred through GHC, lower GI costs constituted an estimated 16% of total annual costs for IBS, 11% for diarrhoea, 11% for constipation and 12.5% for abdominal pain

Components of total costs

Cost categories for Period 3 (the 6 months following the enrolment period) are shown in Table 3. Total out-patient costs exceeded in-patient costs by approximately 5:1. Pharmacy costs were approximately 20% of out-patient costs except for abdominal pain (15%). Radiological tests and other laboratory charges combined were 10–17% of out-patient costs (highest for abdominal pain). Patients with abdominal pain were the most likely to incur costs associated with seeing a specialists, and patients with constipation were most likely to incur costs associated with admission to a hospital.

Table 3.   Components of total costs for Period 3 (mean costs)
CategoryIBSDiarrhoeaConstipationAbdominal PainAll patients
  1. GI, gastrointestinal; IBS, irritable bowel syndrome.

  2. * Out-patient total cost incorporate all subcategories listed below, but not out-of-pocket expenses; † Out-of-pocket total expenses incorporate both non-prescription medications and alternative treatments. Data are for 6 months only.

In-patient4463841050886640
Out-patient total*24832636334939233014
 Primary care office visit619577787689655
 GI clinic office visit401491579727528
 Mental health office visit10982699895
 Pharmacy556514637596571
 Radiology239210265506307
 Emergency80185159203142
 Lab8114684148110
 Other398447769956606
Claims data total29293020439948093654
Out-of-pocket†204147195152178
 Non-prescription100691107589
 Alternative10478857789
All sources total31333167459449613832

Cost impact of symptom severity

For every diagnostic group except constipation, annual mean GHC costs tended to be highest for patients with the most severe symptoms (see Figure 4). The difference between mild and severe symptoms was statistically significant for diarrhoea [F(2,205) = 3.741, P < 0.05] and abdominal pain [F(2,280) = 5.699, P < 0.05].

image

Figure 4.  Total annual costs by severity(mean). □Mild bsl00023Moderate bsl00001Severe. Asterisk indicates that the difference between mild and severe symptom groups was significant for diarrhoea and abdominal pain (P < 0.05). IBS, irritable bowel syndrome.

Download figure to PowerPoint

Satisfaction with treatment

Total costs at follow-up (Periods 3 and Period 4 combined) tended to be higher for patients who recorded ‘satisfactory relief’ of their bowel symptoms compared to those who recorded ‘no satisfactory relief’, with the exception of IBS patients. These differences were not statistically significant, and they were opposite to our expectation of lower costs in patients who achieved satisfactory treatment. See Synergy version for details.

First time consulters vs. returning patients

We hypothesized that health care costs would be higher for first-time consulters compared to returning patients during Period 3 due to the costs of diagnostic tests. We found that the opposite was true (see Figure 5). For the 6-month period following the index visit (Period 3), costs for returning patients were significantly higher for IBS patients [F(2,534) = 5.792, P < 0.05]. For constipation and abdominal pain, these trends were not statistically significant.

image

Figure 5.  Total costs, first time and returning patients, for Period 3 (mean). bsl00023Returning bsl00001First time. Asterisk indicates that for irritable bowel syndrome (IBS), the difference between first time and returning patients was statistically significant (P < 0.05).

Download figure to PowerPoint

To investigate this unexpected finding, post-hoc t-tests were used to investigate whether the difference between first time consulters and returning patients could be explained by differences in IBS symptom severity at enrolment, somatization test scores, number of co-morbid medical conditions, or satisfaction with medical treatment (see Table 4).

Table 4.   Comparison of first time consulters to returning patients on variables that may influence health care costs (mean scores)
Diagnostic groupSatisfactory relief (percent reporting satisfactory relief) Somatization (score: 0–26)Comorbidity (number of comorbid conditions, 0–16) IBS symptom Severity (score: 0–500)
New (%)Returning (%)NewReturningNewReturningNewReturning
  1. IBS, irritable bowel syndrome.

  2. * Differences between new patients (first time consulters for their bowel symptoms) and returning patients (the index visit was not the first time they had sought consultation for their bowel symptoms) are statistically significant (P < 0.05).

Abdominal pain68598.92*11.631.892.74188.82*233.99
Constipation73679.02*10.912.00*3.05185.29*237.01
Diarrhoea81*568.73*11.131.83*2.56180.84*226.93
IBS65*519.72*11.762.47*3.56189.86*266.57
All groups71*569.18*11.532.08*3.13186.26*248.99

Using GLM analysis, we then re-investigated the impact of consulter status on total costs after adjusting for IBS symptom severity at enrolment, somatization, the number of co-morbid medical conditions, or satisfaction with medical treatment. Accounting for these covariates, we found there was no longer a statistically significant difference between new patients and returning patients.

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

Accurate estimates of per-patient health care costs for specific disorders are important in two contexts. First, they are used to make decisions about the allocation of health care resources. Health planners may multiply the per patient cost of care by the known prevalence of the disorder to arrive at an estimate of total direct health care costs and then compare these costs to direct costs for other disorders in deciding how to allocate patient care resources and research funds. Cost data are also used by pharmaceutical companies to decide which diseases to target for drug development and how to price new drugs (e.g., whether development costs are likely to be recovered). The high prevalence of IBS and other functional gastrointestinal disorders (FGIDs) has prompted an interest in their costs to society, individuals and employers.

In their review of the literature, Maxion-Bergemann and colleagues3 found that total annual direct health care cost estimates for IBS patients varied widely across different studies from $348 to $8750 per patient. According to the authors, reasons for this wide range in cost estimates include ‘…features of the investigated patient group (age, limitation to health seekers or all IBS patients, comorbidity, severity of symptoms), database used, method of data collection (retrospective or prospective), varying cost components, time-point of data collection in relation to index-date of IBS diagnosis, method of cost calculation (incidence or prevalence based) and different healthcare systems in the US and the UK’ (pp.21–22).

In our study, in order to arrive at an accurate estimate of direct health care costs, we collected data from a large HMO where costs are calculated as the organization’s actual costs for delivering the care, taking into account all cost categories. This differs from studies which have estimated costs based on billable charges for a particular mix of insurance providers or amounts that Medicare or other government agencies are willing to pay for the services. Our cost estimates are likely to be lower than estimates based on billable amounts or collectible amounts. However, insofar as the large patient population at GHC is representative of patients in other parts of the country and insofar as professional salaries and other costs are comparable to other parts of the United States, our cost estimates may be generalizable to the 70 million adults who receive their health care through HMOs. Additional strengths of our study, which distinguish it from previous studies of health care costs for the functional GI disorders, are: (i) we excluded patients with serious comorbid medical conditions such as IBD, GI malignancy, or malabsorption which are likely to incur high total health care costs, and (ii) we included estimates of out-of-pocket expenses for non-prescription medications and non-traditional health care. We also conducted analyses comparing different FGID patients: IBS, diarrhoea, constipation and abdominal pain.

Total direct health care costs for FGIDs

The mean annual cost for IBS patients in our study was $5054 per patient, including out-of-pocket expenses (see Figure 1). This is within the range reported by Maxion-Bergemann and colleagues. The most directly comparable studies were two previous reports from our research group that estimated direct health care costs for IBS from this same population at Group Health Cooperative. In 2000, Levy et al.11 reported the costs of young adults with IBS who had children between ages seven and 15. For this sample, the average annual cost (not including out-of-pocket expenses) for the period 1993–1995 was $3166. In 2001, Levy et al.12 reported that the total health care cost for IBS was $4044 based on data for 3153 IBS patients during the period 1993–1997. These two previous studies and the current study used the same methodology for classifying health care costs. It is likely that the differences in total health care costs between the first and second studies listed above was related to the age of the patients with IBS: the average age in the 2001 study was 53 years, whereas the 2000 study included younger adults who were selected because they had children aged 7–15 years. The average age of subjects in the current study was 51 years and the cost estimates come from the period 2000–2003.

Health care costs increase over time at a rate that exceeds the rate of inflation in the broader economy. Taking into account the 6-year time lapse between the Levy 2001 study12 and the current study, as well as the inclusion of out-of-pocket expenses, the cost estimates for IBS are comparable ($5054 for the current study vs. $4044 for the Levy 2001 study). The cost estimates for the current study are, we believe, representative of the true direct health care costs for patients with functional GI disorders.

When comparing the four groups of FGID patients in our sample, we expected to find higher total costs for IBS patients compared to patients diagnosed with abdominal pain, constipation, or diarrhoea. However, we found that total costs were significantly lower for IBS compared to abdominal pain and constipation (Figure 1). This may be explained, in part, by the demographics of the IBS patients compared to all other patients in the study: the IBS patients were significantly younger and had a higher average socioeconomic status as judged by the proportion whose partners were college graduates. In separate analyses, we found both younger age and higher education to be associated with lower total GHC health care costs.

Lower GI Costs

For all four diagnostic groups, lower GI costs comprised about 8% of GHC total costs. This is similar to our findings in previous studies: in the Levy 2000 study,11 we reported that 11.5% of the ambulatory care visits made by IBS patients were for GI indications, and in the Levy 2001 study12 we reported that GI-related health care accounted for only 10% of total health care costs for IBS. Therefore, our data confirm that most of the total health care costs paid by insurers for patients with IBS are for non-GI, comorbid conditions. Furthermore, our data show for the first time that health care costs in chronic constipation, functional diarrhoea and functional abdominal pain are also predominantly due to non-GI comorbid conditions.

The out-of-pocket costs in our study were exclusively related to the treatment of bowel symptoms. These out-of-pocket costs ranged from 47% of total lower GI health care costs (GHC claims plus out-of-pocket expenses survey) for IBS to 31% for abdominal pain total lower GI costs (Figure 3).

Components of total costs

An American Gastroenterological Association publication on The Burden of Gastrointestinal Diseases13 estimated $1.35 billion (1998 dollars) in total annual direct costs of IBS, assuming an IBS prevalence rate of 15 percent. In that report, hospital in-patient care accounted for 74% of total direct costs, office visits 17%, drugs 6%, hospital out-patient care 2.5% and hospital emergency services 0.5%. For patients with diarrhoea, the American Gastroenterological Association (AGA) report found that in-patient care accounted for 44% of total costs and out-patient office visits 30%. All of these estimates were based on a review of the literature and included data from a wide variety of health care settings.

In marked contrast to the AGA Burden of Gastrointestinal Disease report, our study of a large HMO showed that in-patient costs comprise a relatively small portion of total GHC costs (17.5% overall and 15.2% for IBS) (see Table 3). Our findings are in line with the proportion of health care costs attributable to in-patient care in both of the previous studies of this population.11, 12 We believe the relatively high proportion of total direct health care attributed by the AGA Burden of Gastrointestinal Diseases report13 to inpatient care is misleading and that the more modest estimates of 15.2% for IBS and 17.5% for FGIDs overall is more accurate and generalizable. However, we are not able to account for the discrepancy between our data and those reported by the ‘Burden of Gastrointestinal Diseases’ report.

In our study, the highest GHC cost categories for out-patient care for all FGIDs combined were professional charges for evaluation and management at office visits (42.4%), pharmacy costs (18.9%) and radiological testing (10.2%). Office-based evaluation and management charges were in three categories: primary care visits (51.3%), GI clinics (41.3%) and mental health services (7.4%). For IBS, the major components of out-patient health care costs were office-based evaluation and management charges (45.5%), pharmacy (22.4%) and radiological services (9.6%). For IBS patients, office-based evaluation and management costs separated into primary care (54.8%), GI clinics (35.5%) and mental health visits (9.7%).

Out-of-pocket expenditures

Total out-of-pocket annual expenditures for non-prescription drugs and alternative medications, extrapolated from patient reports for the last 3 months of Period 3, ranged from $294 for diarrhoea to $406 for IBS. However, despite their relatively low cost compared to total direct costs, alternative medicine and home remedies were used by as much as a third of all FGID patients in our sample. Approximately 56% of IBS patients in our sample reported using alternative medicines or home remedies and approximately 51% used non-prescription drugs.

Method of identifying/recruiting patients influences cost estimates

In this study, we identified patients with FGIDs based on a diagnosis assigned at a clinic visit, and we contacted these patients to complete questionnaires within 2 weeks of this index visit. This method differs from other studies in which patients are identified based on receiving the diagnosis of an FGID at any point within a longer period of time. In our study, we believe the generally higher health care costs seen in the 6-month period following enrollment in the study compared to the preceding 6 months or the subsequent 6 months is due to our having recruited patients at a time in which they were experiencing an exacerbation (or a new onset) of bowel symptoms. This confirms that the methods used to ascertain subjects has an impact on cost estimates and should be considered in interpreting data on cost of care. (The estimates of average annual health care costs in our manuscript are based on the average of two years, including the 12 months prior to the index visit and the 12 months following the index visit.)

Symptom severity and satisfaction with treatment

As hypothesized, total GHC health care costs were highest for patients with the most severe bowel symptoms at the index visit (Figure 4), and differences in costs between patients with mild, moderate and severe symptoms were statistically significant for diarrhoea and abdominal pain. This confirms that the method of recruiting patients (i.e., the inclusion criteria) in cost of care studies has a significant impact on cost of care estimates. The estimates reported in this study are believed to be representative of patients seen in office-based practices because no exclusions were imposed based on disease severity or response to medical treatment.

We anticipated that patients who were satisfied with the outcome of their treatment for bowel symptoms would have lower GHC health care costs than patients with persisting symptoms. However, satisfaction with the outcome of treatment had no significant impact on total costs and, in fact, tended to be associated with higher costs.

First time consulters compared to returning patients

We anticipated that total health care costs would be higher, especially during Period 3, for patients who reported they were consulting physicians about their bowel symptoms for the first time, because we reasoned that health care providers would be more likely to perform endoscopy, radiological exams, and other diagnostic tests. However, we found the opposite (Figure 5); for each diagnostic group except diarrhoea patients, the GHC costs of returning patients were higher compared to first time consulters. In follow-up exploratory analyses, we found that the higher costs for returning patients could be explained by greater IBS symptom severity, more comorbid conditions at enrollment, and reduced satisfaction with the outcome of treatments for returning patients in comparison to first time consulters.

Effects of sex, age, race and education on health care costs

Although some previous reports suggest that racial and ethnic minorities utilize health care less for functional GI disorders and incur lower health care costs,14 there were no race-related differences in health care costs for the members of this prepaid insurance plan. This difference likely reflects economic barriers to obtaining health care in the broader population. We also found no differences between males and females in health care costs provided through GHC; however, women did report using more non-prescription and alternative medicine treatments.

Age, as expected, was linearly related to total health care costs provided through GHC, although we saw no effect of age in the amounts spent on non-prescription and alternative health. The latter was unexpected and may reflect economic disincentives for older subjects who are on limited incomes to purchase these services.

Education had an intriguing, paradoxical effect on health care costs: college educated subjects incurred significantly lower health care costs through their insurance program, but they spent significantly more on non-prescription and alternative medicine treatments. This discrepancy warrants further research to determine whether it may be related to a reduced willingness by college graduates to take time off work to go to a medical clinic and greater economic resources enabling them to take advantage of non-traditional treatments.

Conclusion

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

We have shown that, in an HMO setting where costs are calculated based on the organization’s cost to provide the service, mean total annual direct health care costs incurred by GHC as the insurance provider were $5049 for IBS, $6140 for functional diarrhoea, $7522 for chronic constipation and $7646 for functional abdominal pain. When out-of-pocket expenses are added, the total cost was $5.455 for IBS, $6434 for diarrhoea, $7912 for constipation and $7950 for abdominal pain.

A novel finding of this study is that the proportion of total health care costs attributable to in-patient care was 15.2% for IBS, not 74% as suggested by the AGA ‘Burden of Gastrointestinal Diseases’ report.15 We have identified factors that may explain differences between studies in the estimated cost of care for FGIDs: these include disease symptom severity (associated with increased costs), satisfaction with treatment (associated with increased costs), and whether the patient was recruited into the study in close proximity to an exacerbation of their bowel symptoms (associated with increased costs). An unanticipated finding was that patients consulting for the first time for bowel symptoms do not have higher health care costs than returning patients, and this seems to be related to the fact that returning patients have on average more severe bowel symptoms, a greater number of comorbid conditions, and poorer treatment outcomes.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

Declaration of personal interests: Dr W. E. Whitehead has served as a speaker, a consultant and an advisory board member for Novartis Pharmaceuticals, Takeda Pharmaceuticals, GlaxoSmithKline and McNeil Pharmaceuticals, and has received research funding from Novartis Pharmaceuticals, Takeda Pharmaceuticals and Sandhill Scientific. Declaration of funding interests: This study was supported in full by Novartis Pharmaceuticals Corporation and NIH grants RO1 DK31369 and R24 DK067674.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References
  • 1
    Saito YA, Schoenfeld P, Locke GR3. The epidemiology of irritable bowel syndrome in North America: a systematic review. Am J Gastroenterol 2002; 97: 19105.
  • 2
    Russo MW, Wei JT, Thiny MT, et al. Digestive and liver diseases statistics. Gastroenterology 2004; 126: 144853.
  • 3
    Maxion-Bergemann S, Thielecke F, Abel F, Bergemann R. Costs of irritable bowel syndrome in the UK and US. Pharmacoeconomics 2006; 24: 2137.
  • 4
    Whitehead WE, Levy RL, Von KM, et al. The usual medical care for irritable bowel syndrome. Aliment Pharmacol Ther 2004; 20: 130515.
  • 5
    Saunders KW, Stergachis A, Von Korff M. Group health cooperative of pudget sound. In: StromBL, ed. Pharmacoepidemiology, 2nd edn. New York NY: John Wiley, 1994: 24762.
  • 6
    Drossman DA, Corazziari E, Talley NJ, Thompson WG, Whitehead WE. Rome II modular questionnaire Rome II. The Functional Gastrointestinal Disorders. Diagnosis, Pathophysiology and Treatment: A Multinational Consensus. McLean, VA: Degnon Associates, 2000: 66988.
  • 7
    Francis CY, Morris J, Whorwell PJ. The irritable bowel severity scoring system: a simple method of monitoring irritable bowel syndrome and its progress. Aliment Pharmacol Ther 1997; 11: 395402.
  • 8
    Drossman DA, Patrick DL, Whitehead WE, et al. Further validation of the IBS-QOL: a disease-specific quality-of-life questionnaire. Am J Gastroenterol 2000; 95: 9991007.
    Direct Link:
  • 9
    Derogatis LR. Brief Symptom Inventory (BSI) administration, scoring, and procedures manual (3rd edn.). Minneapolis: NCS Pearson, Inc, 1993.
  • 10
    VanZanten SJOV, Talley NJ, Bytzer P, Klein KB, Whorwell PJ, Zinsmeister AR. Design of treatment trials for the functional gastrointestinal disorders. In: DrossmanDA, CorazziariE, TalleyNJ, ThompsonWG, WhiteheadWE, eds. Rome II: The Functional Gastrointestinal Disorders, 2nd edn. McLean, VA: Degnon Associates, 2000: 577622.
  • 11
    Levy RL, Whitehead WE, Von Korff M, Feld AD. Intergenerational transmission of gastrointestinal illness behavior. Am J Gastroenterol 2000; 95: 4516.
    Direct Link:
  • 12
    Levy RL, Von Korff M, Whitehead WE, et al. Costs of care for irritable bowel syndrome patients in a health maintenance organization. Am J Gastroenterol 2001; 96: 31229.
    Direct Link:
  • 13
    American Gastroenterological Association. The burden of gastrointestinal diseases. Bethesda, MD: American Gastroenterological Association Press, 2001.
  • 14
    Sandler RS. Epidemiology of irritable bowel syndrome in the United States. Gastroenterology. 1990; 99: 40915.
  • 15
    Sandler RS, Everhart JE, Donowitz M, et al. The burden of selected digestive diseases in the United States. Gastroenterol 2002; 122: 150011.