Volume 40, Issue 5 , Pages 696-703, November 2010
The Direct and Indirect Costs of Opioid-Induced Constipation
Article Outline
Abstract
Context
Treatment with strong opioids is connected with frequent and problematic side effects. One of the most common side effects is opioid-induced constipation (OIC). The discomfort of OIC can limit the effectiveness of pain therapy. Because constipation typically persists for as long as opioid therapy is administered, its effects on the quality of life (QoL) of patients need to be taken seriously. Data and published studies on the cost implications of OIC are, however, scarce.
Objectives
To estimate the direct and indirect costs of OIC in a defined patient population during treatment with strong opioids.
Methods
The study is based on patient data from a Swedish noninterventional study, UPPSIKT. The cost analysis is based on 197 patients treated with strong opioids over a six-month period. Direct and indirect costs in this article are calculated per patient-month, and the cost for OIC is estimated as the difference in mean costs between months with and without constipation.
Results
The total costs per patient-month for patients with severe constipation are significantly higher than those for patients with mild, moderate, or no constipation. Patients with severe constipation have the highest total costs, Euro (EUR) 1525 per patient-month, whereas patients with mild, moderate, and no problems cost EUR 1196, EUR 1088, and EUR 1034, respectively.
Conclusion
Opioid use is costly to society, and the costs vary with OIC severity. OIC is discomforting, affects the QoL of patients, and can limit an effective pain therapy.
Key Words: Opioid-induced constipation, OIC, costs, Sweden
Introduction
Chronic pain is a major health care problem, affecting almost 20% of adult Europeans.1, 2 The International Association for the Study of Pain defines pain as “an unpleasant sensory and emotional experience associated with actual or potential tissue damage, or described in terms of such damage.”3 The most common types are back injuries, headaches, and joint pain. Chronic pain also can be caused by diseases or disorders, such as cancer.4
Strong opioids, including morphine, codeine, oxycodone, methadone, and fentanyl, are commonly prescribed for the relief of chronic pain.4, 5 Daily problems with common side effects of opioid administration include sedation, dizziness, nausea, vomiting, constipation, and respiratory depression.6 Most patients develop tolerance against side effects, but constipation is a major problem that may not diminish over time.7
As constipation is the most frequent side effect associated with long-term opioid therapy, a thorough history of the patient’s bowel habits should be taken before initiating opioids. Prophylaxis should be considered when starting opioid therapy. Common constipation prophylaxis (e.g., fiber, fluids, and exercise) may not be sufficient, and oral laxatives are routinely needed.8
The discomfort of opioid-induced constipation (OIC) can limit the effectiveness of pain therapy via reduced compliance.9, 10 Studies have shown that opioid-treated patients with constipation have higher health care utilization, are absent from work more often, and report impairment in performing their daily activities compared with nonconstipated patients.11, 12 In opioid-treated cancer patients, constipation significantly impacts opioid use patterns, resource utilization, and costs. Alleviation of constipation may, therefore, optimize opioid therapy and reduce costs.13 OIC also has a serious impact on quality of life (QoL).14, 15, 16 Because constipation may persist for as long as opioid therapy is administered, it may be costly and debilitating and justifies being taken seriously. To our knowledge, however, the actual cost implications of OIC are poorly understood. The objective of this study was to estimate the direct and indirect costs of OIC in a naturalistic cohort of patients treated with strong opioids in Sweden.
Methods
Approach
The costs of treatment were calculated by multiplying resource use data from a large sample of patients treated with strong opioids by unit costs. The direct costs of OIC were calculated as observed resource usage multiplied by the corresponding unit costs for each patient during each study month. Indirect costs were calculated as the value of foregone productivity because of illness, using the human capital approach.
Patient-reported constipation was used to split the data set into two categories: patient-months in which patients had constipation and patient-months in which they did not. The costs of OIC were calculated as the mean differences between patient-months with and without constipation.
The Study Population
We used resource use data from UPPSIKT, a noninterventional study in Sweden. UPPSIKT includes data on patients treated with strong opioids (all types and administration forms) during a six-month period. Sample inclusion was based on physician-determined long-term need (minimum six months) of strong opioids, older than 18 years, and literacy in Swedish, which ensured generally a representative sample of patients treated with strong opioids in Sweden. Depending on the underlying disease, patients were treated by chronic pain specialists, palliative care specialists, cancer specialists, or primary care physicians. Each patient completed an informed consent form before inclusion in the study. The study was approved by the appropriate independent ethics committee.
The study enrolled 331 patients from 38 clinics with geographical dispersion in Sweden including both patients with malignant and nonmalignant pain. Data were collected via a baseline questionnaire completed by a physician and six monthly questionnaires completed by the patients over the following six months. A letter was attached to each questionnaire highlighting the importance of responding to all questions even if there were no changes in the patient’s health status since their last completed questionnaire.
The baseline questionnaire captured information about age, gender, cause and type of pain, pharmacological treatment, side effects, reason for initiating or change of current strong opioid, and assessment of patients’ well-being because of their pain.
The six patient questionnaires solicited information about pain relief, health care contacts, pharmacological treatment, reason for any change in pain treatment, side effects, QoL, and ability to work. Patients were instructed to tell if they had experienced problems with constipation and rate that on a scale from 0 to 10, where 0 represented no constipation and 10 represented extensive problems with constipation. In this cost study, patients were then divided into three groups: mild (score 1–3) moderate (score 4–7) and severe (score 8–10). In the questionnaires, the patients were asked to estimate their QoL on a scale from 0 to 10, where 0 represented the worst possible QoL and 10 represented the best possible QoL. Patients also were asked to indicate each month to what extent they were satisfied with their pain treatment on a scale from 0 to 10, where 0 equals very dissatisfied and 10 equals very satisfied.
Costing of Resource Use
Costs consist of direct and indirect costs. The direct costs consist of the value of resources used for the treatment of patients, that is, costs for inpatient care, outpatient care, and pharmaceuticals. These costs are valued according to market prices. The indirect costs reflect the value of lost production because of sick leave. There are different techniques for estimating these; but the standard methodology in Sweden and in most other countries, and the method we used, is the human capital approach, which uses lost earnings as a substitute for the impact that premature death and disability have on individuals and society.17, 18
Costs in this article are calculated per patient-month, with and without constipation, and the costs for constipation are thus estimated as the difference between the total costs for patient-months with and without constipation. Costs are expressed in Euros (EUR), 2008 prices.
Health care resources used for treatment of patients include outpatient care, telephone consultations and visits to specialists and physicians in primary care, telephone consultations and visits to nurses, visits to physiotherapists, and inpatient care. All costs for health care resources are based on official price lists from neurological and pain clinics in the Swedish health care regions (Table 1).19, 20, 21, 22, 23, 24, 25
Table 1. Direct and Indirect Cost Units (EUR, 2008 Prices)
| Costs in EUR | |
|---|---|
| Direct costs | |
| Health care visits | |
 Physician, specialist (hospital) | 417 |
| Physician, primary care | 160 |
| Nurse | 55 |
| Physiotherapist | 55 |
| Telephone consultationsa | |
| Physician, specialist (hospital) | 208 |
| Physician, primary care | 80 |
| Nurse | 27 |
| Physiotherapist | 27 |
| Inpatient day | 432 |
| Indirect costs | |
| Production loss, one day | 199 |
aThe cost for a telephone consultation is assumed to be half the cost of a visit. |
Costs for pharmaceuticals were collected from Farmaceutiska specialiteter i Sverige (FASS) 2008 (Pharmaceutical Specialties in Sweden).26 The drug costs included in the analysis were costs for opioids, other analgesics, and laxatives.
The production loss was estimated using the average yearly income from work in Sweden, adjusting for payroll taxes by multiplying by 1.41 (the employment payroll taxes for the Swedish population 20–64 years is 41%),27, 28 and assuming an average of 250 working days in a year.
Statistical Analysis
Statistical analyses, with descriptive statistics and subgroup analysis of resource utilization related to disease severity, were performed. Statistical tests performed on the data set were t-tests and F-tests to test significant differences between two or more groups of patients. The type of test is clearly indicated in the Results section. All statistical tests were performed at the 0.05 level of significance and were two-sided. SD refers to standard deviation.
Results
Patient Characteristics
A total of 197 patients (i.e., 60% of all the patients enrolled) responded to all six monthly questionnaires in the UPPSIKT study. Of them, the majority, 117 patients, were women (59%). The respondents’ age ranged between 23 and 95 years, with an average age of 61 years (SD=14.0). One hundred twenty-eight of these patients were retirees, 46 reported to be working, and the rest were assumed to be either unemployed or students. Noncancer-related pain (80%) was the most common cause of pain; 15% of the patients reported cancer-related pain and 5% percent reported a combined cause of pain. The types of pain the patients were experiencing were nociceptive (85%), neuropathic (55%), and visceral (11%). Reports of more than one type of pain were common (Table 2).
Table 2. Patient Characteristics
| Variables (n=197) | All Patients |
|---|---|
| Male, n (%) | 80 (41) |
| Female, n (%) | 117 (59) |
| Mean age, years (range) | 61 (23–95) |
| Reported cause of pain, n (%) | |
| Cancer-related | 29 (15) |
| Noncancer-related | 158 (80) |
| Combination | 9 (5) |
| Time treated with strong opioids, n (%) | |
| Opioid naive | 13 (7) |
| <1 month | 9 (5) |
| 1–6 months | 31 (16) |
| 6 months–1 year | 29 (15) |
| 1–5 years | 72 (36) |
| >5 years | 42 (21) |
Pain Treatment
At baseline, more than 20% of the patients had been treated with strong opioids for more than five years. Almost 40% of patients had been taking opioids between one to five years, and approximately 15% had been taking opioids for between six months to one year and for one month to six months. Around 5% had been treated with opioids for less than one month, whereas 7% of the patients were opioid naive.
The most commonly used strong opioids were oxycodone, morphine, methadone, and buprenorphine. They accounted for 34%, 16%, 14%, and 14%, respectively, of all opioid use during the six months.
As shown in Table 3, the most common type of outpatient care was specialist visit (0.29 visits per patient-month) followed by nurse visit (0.19 visits per patient-month). There were on average 0.32 telephone consultations with specialists per patient-month and 0.19 telephone consultations with nurses on average per patient-month. Inpatient care was rare (two to nine patients per month). The mean number of inpatient days per patient-month in UPPSIKT was 0.19 (SD=0.85).
Table 3. Mean Number of Visits or Telephone Calls in Outpatient Care per Patient-Month in the UPPSIKT Study (n=197)
| Type of Outpatient Care | Mean | SD |
|---|---|---|
| Health care visits | ||
| Physician, specialist (hospital) | 0.29 | 0.44 |
| Physician, primary care | 0.08 | 0.20 |
| Nurse | 0.19 | 0.67 |
| Physiotherapist | 0.02 | 0.11 |
| Other | 0.03 | 0.13 |
| Telephone consultations | ||
| Physician, specialist | 0.32 | 0.46 |
| Physician, primary care | 0.06 | 0.22 |
| Nurse | 0.19 | 0.44 |
| Physiotherapist | 0.02 | 0.11 |
| Other | 0.03 | 0.13 |
Opioid-Induced Constipation
Around 60%–70% of the patients reported that they experienced some degree of a problem with constipation each month. During the study, approximately 12% of patient-months were categorized as months with severe problems with OIC, 25% as moderate, 26% as mild, and 37% as months with no constipation.
The most commonly used laxatives in the UPPSIKT were sodium picosulphate, macrogol, sodium chloride, sodium bicarbonate, and lactulose. These agents accounted for 29%, 21%, and 21%, respectively, of all laxative use in the study.
Health Care Resource Use
The calculated costs, illustrated in Fig. 1, are based on the total number of patient-months with mild (302 months), moderate (288 months), or severe (135 months) constipation, and the total number of months without constipation (419 months). (Note that this does not add up to 1182 months [6×197] because of missing data for some patient-months regarding the degree of constipation.)
Fig. 1
Mean cost (EUR) per patient-month with no, mild, moderate, and severe constipation (SD within parentheses).
Mean total costs were EUR 1034 (SD=1350) for months with no reported constipation. The largest cost item was indirect costs because of sick leave, EUR 596 per patient-month, followed by outpatient costs, EUR 196 per patient-month.
Patients with severe constipation had the highest total costs EUR 1525 (SD=1711) per patient-month, whereas patients with mild and moderate problems had EUR 1196 (SD=1544) and EUR 1088 (SD=1489), respectively. The largest cost item in all three groups with constipation was indirect costs followed by costs for outpatient care.
Patients in months with severe problems had significantly higher total costs than patients in months with no (t=3.43; P=0.001), mild (t=1.99; P=0.047), and moderate (t=2.68; P=0.008) problems with OIC. The indirect costs for patients experiencing severe constipation differ from costs of moderately constipated and not constipated patients (t=2.10; P=0.036 and t=2.46; P=0.014, respectively). The direct costs for patients with severe problems differ from costs of patients with no and mild constipation (t=2.80; P=0.005 and t=4.05; P=0.000, respectively). At a disaggregated level, there was no significant difference in inpatient care between patients with severe constipation and other patients. The outpatient care costs for patients experiencing severe constipation differed significantly from patients with mild (t=3.98; P=0.000) or no constipation (t=4.73; P=0.000). The costs for drugs differed significantly between patients with severe constipation and patients with no (t=3.17; P=0.002), mild (t=5.11; P=0.000), or moderate (t=3.44; P=0.001) problems with OIC.
Patients in the four categories (no, mild, moderate, or severe constipation) were divided into subgroups depending on length of opioid use to test for differences within the categories. However, we found no significant differences in costs depending on length of opioid use. For other subgroups, such as cause of pain, no significant difference in costs was detected. This is probably because of the fact that the sample is too small for such analysis.
The indirect cost calculations are based on the assumption that everyone working (46 patients) was employed full time. If we assume instead that everyone working instead works only 75%, the cost of patients experiencing severe constipation decreases to EUR 1301 and indirect costs decrease to about EUR 673. Mean indirect costs for those with no and moderate constipation decrease to about EUR 156 and for those with mild constipation to about EUR 208. The new difference in total costs between patients with no constipation and those with severe constipation was approximately EUR 365 vs. EUR 417 under the assumption that everyone working was employed full time.
QoL and Satisfaction with Pain Treatment
As illustrated in Fig. 2, the mean QoL per month and satisfaction with pain treatment varied by patient-month. Patients with no, mild, moderate, and severe constipation reported QoL scores of 4.9, 4.9, 4.7, and 3.8, respectively. There was a significant difference for QoL between patients depending on degree of problem with OIC (F=10.26; P=0.000), and when testing for difference in QoL between patients reporting severe problems with OIC and the other categories, we also found a significant difference (F=−5.42; P=0.000). Moreover, there was a significant difference between the categories for satisfaction with pain treatment (F=13.82; P=0.000). Likewise, patients with severe constipation seemed to be less satisfied with their pain treatment compared with the other patient groups (t=6.04; P=0.000). The mean reported satisfaction with pain treatment per patient-month with no, mild, moderate, and severe constipation was 6.6, 6.6, 6.2, and 5.2, respectively.
Fig. 2
Reported estimation of QoL and reported satisfaction with pain treatment, mean for all six months (0
=worst possible and 10=best possible).
Discussion
This study estimated the costs of all major resources associated with the treatment of pain with strong opioids in Sweden using real patient data. Based on the Swedish noninterventional study UPPSIKT, we conclude that the mean costs per patient-month with severe problems with constipation are significantly higher than for patients with mild, moderate, or no constipation. The difference per patient-month for patients with severe constipation compared with patients with no constipation is EUR 490 for total costs and EUR 188 for direct costs. Specifically, costs for outpatient care and drug contribute the most to the difference in direct costs. According to this study, costs for OIC do not differ depending on the length of opioid use.
One strength of the study is that strong opioid treatment duration among the patients varies considerably, that is, from being opioid naive to taking opioids regularly for more than five years. Almost 60% were treated for more than one year, and of those, approximately 40% were treated for more than five years. Another strength of the UPPSIKT study is the relatively long follow-up time of six months. Moreover, data on both costs and constipation severity were collected on a monthly basis, which reduces the risk of recall error.
Indirect costs because of sick leave are the largest cost item in all the different groups of severity although this should be interpreted with caution. A weakness of the study is that the patients’ mean hours worked are missing. Therefore, the estimation of lost production in this study assumes that all patients being employed full time and the costs of production loss may thus be overestimated. To find out how this assumption affects the results, a sensitivity analysis was conducted based on the assumption of part-time employment, that is, 75%, which resulted in a reduction in the mean indirect costs per patient-month to EUR 156 from EUR 208.
Patients with severe constipation had lower QoL, as these patients on average responded with lower QoL scores compared with the other patients in the study. This result is in line with other studies showing that OIC can have a serious negative impact on QoL and the daily activities that patients feel able to perform.10, 15, 16 Even if symptomatic therapy, such as laxatives, can provide some relief for patients with constipation, there is clearly a need for new therapies that can reduce or eliminate problems with constipation for patients on pain treatment.16
There is a trade-off between effective pain therapy and the risk of discontinuation because of OIC. To reduce the discomfort because of OIC, the patients might not take the same amount of opioids as they should in the absence of OIC and, therefore, may not receive maximum pain relief. These facts have recently been shown in a patient survey, in which one-third of the patients reported that they had missed, decreased, or stopped using opioids to reduce their problems with constipation.9 Of these patients, the majority experienced increased pain after having decreased or discontinued the opioid use. Because the design of our study did not control for the effectiveness of pain therapy, we may have underestimated the OIC costs.
An assumption made in this study is that patients taking opioids and being constipated have OIC. However, because of differences in age, lifestyle, and underlying disease, there may be other causes of constipation as well. Furthermore, because it is likely that problems with OIC differ depending on the length of opioid use, it would be useful to evaluate the resource use and costs for a homogeneous patient population in a future study.
Conclusions
The results from our cost analysis indicate that OIC imposes substantial costs to society, especially for patients with severe problems with constipation. To our knowledge, no cost estimates of OIC have been published before this study. Patients with severe OIC experience lower life quality than patients with less problems with OIC, which is a result in line with other studies. There is a large unmet need for new therapies that can reduce or eliminate problems with constipation for patients on pain treatment.
Acknowledgments
The study was financed by Mundipharma AB, Sweden. Anna-Carin Berggren is an employee of Mundipharma AB, Sweden.
References
- . Survey of chronic pain in Europe: prevalence, impact on daily life, and treatment. Eur J Pain. 2006;10:287–333
- . Report: Pain in Europe/Den svenska undersökningen The Swedish survey. Göteborg, Sweden: The Swedish Pain Association; 2003;
- . Logic, truth and language in concepts of pain. Qual Life Res. 1994;3:69–76
- . The role of opioids in cancer pain. BMJ. 2005;331:825–829
- . Opioids in chronic non-cancer pain: systematic review of efficacy and safety. Pain. 2004;112:372–380
- Opioid complications and side effects. Pain Physician. 2008;11(2 Suppl):S105–S120
- Meeting the challenges of opioid-induced constipation in chronic pain management—a novel approach. Pharmacology. 2009;83:10–17
- . Management of opioid-induced gastrointestinal effects: treatment. Pharmacotherapy. 2002;22:240–250
- The prevalence, severity, and impact of opioid-induced bowel dysfunction: results of a US and European patient survey (PROBE 1). Pain Med. 2009;10:35–42
- . Incidence, prevalence, and management of opioid bowel dysfunction. Am J Surg. 2001;182(Suppl):11S–18S
- . Opioid antagonists: a review of their role in palliative care, focusing on use in opioid-related constipation. J Pain Symptom Manage. 2002;24:71–90
- . Impact of opioid-induced constipation on patients and healthcare resource use. Eur J Pain. 2006;10(Suppl S1):S172
- . Impact of constipation on opioid use patterns, health care resource utilization, and costs in cancer patients on opioid therapy. J Pain Palliat Care Pharmacother. 2009;23:231–241
- . Impact on quality of life of constipation-associated GI symptoms related to opioid treatment in chronic pain patients: PAC-QOL results from the opioid survey. [abstract 883] J Pain. 2007;8:71
- . Quality of life in patients with different types of functional constipation. Scand J Gastroenterol. 1997;32:1083–1089
- . Opioid-induced bowel dysfunction: prevalence, pathophysiology and burden. Int J Clin Pract. 2007;61:1181–1187
- . The role and estimation of productivity costs in economic evaluation. In: Drummond M, McGuire A editor. Economic evaluation in health care: Merging theory with practice. Oxford, UK: Oxford University Press; 2001;p. 94–112
- . Methods for the economic evaluation of health care programmes. Oxford, UK: Oxford University Press; 2005;
- University Hospital Health Care of Norrland. Pricelist 2008. 2008. Available from http://www.norrlandstingen.se/2008_bil_2_nusprislista_utg3.xls.
- Uppsala County Primary Care. Pricelist primary care 2008. Landstinget i Uppsala, län (County council of Uppsala), Uppsala, 2007.
- Stockholm-Gotland Health Care Region. Pricelist 2008. Available from http://extern.bkv.sll.se/samv.sthlm-gotland/Prislistor/2008/prislistor_2008.htm.
- Southern Regional Health Care Board. Prices and reimbursements for the Southern Health Care Region, 2008. Available from http://www.skane.se/upload/Webbplatser/Sodra%20regionvardsnamnden/prislista/2008/helaprislistan2008.pdf.
- Southeastern Health Care Region. Prices and reimbursements for the Southeastern Health Care Region 2008. 2007. Available from http://www.lio.se/upload/16047/20071218Prislista2008totversion2.pdf.
- Western Health Care Region. Pricelist 2008. 2008. Available from http://www.vgregion.se/upload/svn%20-%20samverkansn%C3%A4mden/Utoml%C3%A4nsprislista%202008.pdf.
- Örebro County Council. Pricelist for primary care 2008. 2008. Available from Örebro läns landsting (County council of Örebro), Örebro.
- FASS (Pharmaceuticals Specialities in Sweden). Stockholm, Sweden: the Swedish Association of the Pharmaceutical Industry AB (LIF), 2008.
- . What is the info available at this website?. Available from www.scb.se2008;Accessed November 2009
- Ekonomifakta. Payroll tax. 2008. Available from http://www.ekonomifakta.se/sv/Fakta/Arbetsmarknad/Arbetsgivaravgift/Sociala-avgifter/.
PII: S0885-3924(10)00463-X
doi:10.1016/j.jpainsymman.2010.02.019
© 2010 U.S. Cancer Pain Relief Committee. Published by Elsevier Inc. All rights reserved.
Volume 40, Issue 5 , Pages 696-703, November 2010
