|Home | About | Journals | Submit | Contact Us | Français|
Hepatitis C virus (HCV) infection is estimated to affect 2% of the general U.S. population and chronic pain is a common comorbidity among persons with HCV. The primary purpose of this study was to compare health service utilization of U.S. military veterans with HCV with and without the presence of comorbid chronic pain.
Cross-sectional study with retrospective review of patient medical records.
One hundred seventy-one U.S. military veterans with confirmed HCV, recruited through a single U.S. Veterans Administration hospital.
Medical service utilization data from the past five years were extracted from participants’ electronic medical records.
Sixty-four percent of veterans with HCV (n = 110) had chronic pain. Veterans with HCV and chronic pain utilized more health services including total inpatient stays (OR = 2.58 [1.46, 4.56]) and days hospitalized for psychiatric services (OR = 5.50 [3.37, 8.99]), compared to participants with HCV and no chronic pain, after statistically adjusting for demographic, psychiatric, substance use, medical comorbidity, and disability covariates. In addition, those with HCV and chronic pain had more total outpatient visits with primary care providers (OR = 1.73 [1.15, 2.59]), physical therapists (OR = 9.57 [4.79, 19.11]), and occupational therapists (OR = 2.72 [1.00, 7.48]).
Patients with HCV and chronic pain utilize medical services to a greater extent than patients with HCV but no chronic pain. Future studies that examine the efficacy of both pharmacological and nonpharmacological pain treatment for patients with comorbid HCV and chronic pain appear warranted.
Hepatitis C virus (HCV) is a common blood-borne infection that affects nearly 2% of the general U.S. population (1–2). Among U.S. military veterans, prevalence of HCV is estimated to be as high as 5.4%–6.6% (3). HCV is a primary cause of cirrhosis, hepatocellular carcinoma, and liver transplantation (4). The prevalence of chronic pain in patients with HCV is higher than in the general medical population, with 67%–83% of patients with HCV having documented chronic pain diagnoses in their medical records (5–8) compared to approximately one-third of patients seen in a primary care setting (9–10). Chronic pain prevalence in patients with HCV is also higher than in patients with diseased livers due to non-HCV-related causes such as hepatitis B virus (5).
Although the etiologies of chronic pain in patients with HCV have not been elucidated, one possible explanation is side effects of interferon therapy, which include pain for some treated patients (11–12). Among patients with HCV who are interferon treatment-naïve, another possible mechanism for increased chronic pain is the upregulation of pro-inflammatory cytokines (13–14), which are proteins that help modulate the body’s response to infection and other diseases. The presence of excess cytokines in the body has been associated with various chronic pain conditions such as neuropathy, rheumatoid arthritis, and Crohn’s disease, among others (15). In addition, comorbid substance use disorders (SUD) and psychiatric distress, which occur at high rates among patients with HCV, have been shown to be related to chronic pain (6, 16–21).
Chronic pain carries with it significant costs in the form of pain-related disability, decreased quality of life, lost work productivity, and costs to the healthcare system as a result of increased medical service utilization (22). Indeed, people who experience chronic pain utilize both outpatient (22–23) and inpatient medical services (23) to a greater extent than patients receiving general medical care who do not experience chronic pain, and this disparity appears to be most pronounced when comparing patients without chronic pain to patients with chronic pain who report the greatest pain-related disability (23). Patients with HCV are also frequent users of medical services and experience significant disability and lost work productivity due to HCV, regardless of chronic pain diagnoses (24–26). High rates of SUD and mental illness in this population (18) may also lead to increased medical and allied health service utilization. However, the extent to which chronic pain in HCV patients increases healthcare service utilization beyond HCV alone is unclear.
The purpose of this study was to compare health service utilization between patients with HCV and chronic pain versus those with HCV and no chronic pain. We hypothesized that health service utilization, both inpatient and outpatient, would be greater in patients with chronic pain versus those without.
Participants were part of a larger study that examined issues of pain and substance use among patients with HCV. Participants were recruited by means of posted advertisements in the United States Veterans Administration (VA) medical facility where the study took place, letters sent to patients who had pending appointments in primary care, announcements made in mental health classes, and referrals from patients being seen by a VA hepatology provider. All interested patients underwent a brief screening to determine eligibility. Eligible participants for the current study were at least 18 years old, English speaking, and had data in their medical record indicating the presence of HCV confirmed by polymerase chain reaction tests. Exclusion criteria included age over 70, diagnosis of hepatocellular carcinoma, advanced liver disease (i.e., Stage 4, Grade 4 on liver biopsy or decompensated cirrhosis), prior liver transplantation, current suicidal ideation, current untreated psychotic-spectrum disorder (e.g., schizophrenia) or bipolar disorder, or any pending litigation or disability compensation for pain. Patients with a history of interferon/ribavirin treatment for HCV were also excluded because treatment has been associated with severe side effects including chronic pain (11–12). Moreover a secondary goal of the parent study was to identify psychosocial and medical factors, such as chronic pain, in treatment-naïve patients with HCV that could be addressed to improve patients’ candidacy for future HCV treatment.
Two hundred eighty-four patients with HCV were screened, of which 171 patients met full eligibility criteria. Individuals with HCV were excluded for history of interferon and/or ribavirin therapy (n = 37), currently seeking disability for pain (n = 28), diagnosis of cirrhosis (n = 11), non-veteran (n = 3), participant in another study (n = 3), untreated psychotic disorder (n = 2), current suicidal ideation (n = 2), history of hepatocellular carcinoma (n = 2), cognitive impairment that precluded participation (n = 2), age over 70 (n = 1), and incomplete responses to eligibility screening questions (n = 3). An additional five patients declined to participate after screening, and 14 were lost to follow-up. All participants signed informed consent to participate, completed a clinical interview and a set of self-report questionnaires (described below), and received a $30 store gift card as compensation. This study was approved by the Institutional Review Board of the VA Medical Center.
Diagnostic and medical utilization data were extracted from the electronic medical record using the Veterans Integrated Service Network-20 (VISN-20) Data Warehouse. The VISN-20 Data Warehouse extracts data from the clinical records of regional VA facilities and two national VA databases. Diagnoses of commonly occurring pain syndromes were obtained using ICD-9-CM codes listed in medical encounter records for the five years prior to the study assessment: chronic neck or joint pain (ICD-9-CM: 716-719.99, 723-723.99, 729-729.99), chronic low back pain (722-722.99, 724-724.99), rheumatism/arthritis (712-712.99, 714-716.99, 720-720.99, 729-729.99), migraine headaches (346.9, 784.0), neuropathic pain (337.0-337.19, 356-357.99, 377-377-99), fibromyalgia (729.1-729.19), and inflammatory bowel disease (558.9).
In addition, VA medical service utilization over the past five years was assessed using the following variables: number of inpatient stays, number of days hospitalized for general medical reasons, number of days hospitalized for psychiatric care, number of combined scheduled and unscheduled outpatient primary care visits, number of outpatient visits for physical therapy, number of outpatient visits for occupational therapy, and number of outpatient visits to a chronic pain specialty clinic. Inpatient hospitalization and outpatient diagnoses for a variety of medical conditions were used to calculate a Charlson Comorbidity Index score (27), which is a measure of global illness severity (28–29). VA service connected disability status, a measure of disability attributable to military service, was coded as service connected versus non-service connected. Aminotransferase to platelet ratio index (APRI) was obtained from the medical record to assess for liver disease severity (30). APRI is a non-invasive diagnostic test for liver disease severity that employs serum laboratory markers and may be used as an alternative to liver biopsy (30). These variables were obtained to control for medical disease comorbidity and liver disease severity in multivariate analyses assessing medical service utilization between patients with and without chronic pain.
Psychosocial questionnaires assessed participant demographic characteristics including age, gender, race/ethnicity, years of education, annual income, and marital status. Age was treated as a continuous variable, while all others were treated as categorical variables with the following levels: gender (male vs. female), race/ethnicity (Caucasian vs. African-American vs. Latino/a vs. Native-American vs. Mixed or Other race), years of education (≤ 12 years vs. > 12 years), annual income (< $15,000 vs. ≥ $15,000), and marital status (single vs. married vs. divorced/separated vs. widowed). Participants were classified as having chronic pain if they: (1) responded affirmatively to a question that asked, “Have you ever been diagnosed or treated for a chronic pain condition?”, (2) responded affirmatively to a follow-up question that asked, “If yes, do you still experience pain from this condition?”, and (3) had at least one chronic pain diagnosis in the past five years from the VA medical record of the facility where the study took place. Participants who reported having a chronic pain diagnosis also indicated the year they received this diagnosis.
To assess quality-of-life, participants completed the Veterans RAND 12-item Health Survey (VR-12), a validated measure of mental and physical health (31). The VR-12 was adapted from the Veterans RAND 36-item Health Survey, which was developed from the MOS RAND SF-36 Version 1.0 (32). Standardized scoring procedures were used to calculate Physical Health and Mental Health component summary scores for participants with complete VR-12 data. Each summary score is calculated as a t-score, with a mean of 50 and standard deviation of 10 among the population of U.S. military veterans. Higher scores indicate better self-reported functioning.
The presence of current major depressive disorder, as well as current and lifetime histories of alcohol and substance use disorder, were assessed by the Structured Clinical Interview for DSM-IV (SCID ), which has demonstrated excellent psychometric properties (34). To reduce participant burden, we limited the number of specific disorders assessed. We chose major depressive disorder, given prior literature indicating its strong association with chronic pain (16), and alcohol/substance use disorders due to their relationship with hepatitis C and chronic pain (6, 18). All SCID interviews were conducted by research assistants who received extensive training by a licensed psychologist in diagnostic clinical interviewing and the SCID assessment tool. Regular supervision of SCID interviewers was conducted by the study’s principal investigator (B.J.M.) to reduce likelihood of coder drift.
Post-traumatic stress disorder (PTSD) was also assessed in the current study. Participants responded to the following question: “Have you ever witnessed or been involved in a terrible experience such as a fire, flood, or bad traffic accident, or seen someone badly injured or killed or feared that your life, safety or security was being threatened?” Participants subsequently completed the PTSD Checklist, Civilian Version (PCL-C; 35). This measure evaluates the extent to which respondents experienced each of 17 PTSD-related symptoms in the past month using a 1–5 numeric rating scale. Items are summed, and higher scores indicate more severe PTSD-related symptoms. Individuals were classified as meeting PTSD criteria if they responded affirmatively to the trauma question and scored at least 50 on the PCL-C. Internal consistency for this measure in the current study was excellent (Cronbach’s alpha = 0.95).
Participants’ perception of pain severity and the extent to which pain interferes with their lives was assessed with the well-validated West Haven-Yale Multidimensional Pain Inventory (WHYMPI ) Severity and Interference scales, respectively. The WHYMPI Severity scale consists of three questions, while the Interference scale consists of 11 questions. Respondents rate their pain experience on a 6-point numeric rating scale, and items for each scale are averaged, with higher scores indicating greater pain severity or interference with daily activities. Cronbach’s alpha for the WHYMPI scales in the current study were as follows: Severity = 0.71, Interference = 0.91. Although we obtained data for other scales of the WHYMPI, pain severity and interference are constructs commonly evaluated in pain research and in the clinical assessment of pain, have been recommended as outcomes in clinical trials by the Initiative on Methods, Measures, and Pain Assessment in Clinical Trials (IMMPACT; 37), and these were the WHYMPI variables of interest for the current study.
The TimeLine Follow-Back (TLFB ) was used to assess self-reported use of prescription opioids in the 30 days prior to the study assessment. The TLFB has been shown to be a valid, reliable, and standardized tool that uses calendar prompts to track the frequency of use of alcohol, illicit substances, and prescription medications (39).
Finally, participants were asked a series of questions about whether they used each of 10 medications and nonpharmacological treatments for managing chronic pain at any time in the past. These included: over-the-counter analgesics (e.g., ibuprofen, Motrin), opioid medications, physical therapy, transcutaneous electrical nerve simulation (TENS), massage, chiropractic treatment, acupuncture, herbal medicines, relaxation/stress management techniques, and group pain management classes. Past chronic pain treatment use questions did not ask participants to specify the type of facility where treatments were received, dates or duration, length of time since treatment was received, or current use of treatments, nor did they assess patients’ perceived efficacy of or satisfaction with past treatments. Questions employed dichotomous response options of “yes” or “no.”
Descriptive statistics were calculated for demographic, quality of life, psychiatric, and substance use variables, as well as pain severity, pain interference, Charlson Comorbidity Index, VA service connected disability, and APRI. Comparisons across all variables were conducted between participants with chronic pain and those without using χ2 tests for categorical variables and independent samples t-tests for continuous variables. Descriptive statistics identified the proportion of participants reporting chronic pain that had tried various pain treatments in the past.
We employed negative binomial regression to examine differences between patients with and without chronic pain in the following medical service utilization outcome variables for the past five years: number of inpatient stays, number of days hospitalized for general medical procedures, number of days hospitalized for psychiatric reasons, number of primary care visits, number of outpatient physical therapy visits, number of outpatient occupational therapy visits, and number of outpatient visits for specialty pain treatment. Negative binomial regression analyzes count outcome data and is used as an alternative to Poisson regression when overdispersion is present (i.e., the assumption of an equal mean and variance in the Poisson model is violated). All logistic and negative binomial models controlled for participant age, gender, Charlson Comorbidity Index, VA service connected disability, APRI, diagnosis of current major depressive disorder, current PTSD, current alcohol use disorder, current substance use disorder, and other demographic characteristics found to significantly differ between participants with and without chronic pain in bivariate analyses. An α-level of p < 0.05 was used for all inferential analyses.
Of the 171 participants with HCV, chronic pain was reported by 110 participants (64%). Six individuals did not have a chronic pain diagnosis in their medical record, indicating that pain had either not been a focus of their medical treatment in the past five years or they had sought treatment for pain outside the VA medical facility where the study took place. The 104 participants with current chronic pain did not significantly differ from the six participants with past chronic pain on any quality of life, psychiatric, pain severity, pain interference, substance use, or demographic variable, or on measures of medical comorbidity, disability, and liver disease severity (all ps > 0.10). All statistical analyses were performed on the full sample of N = 171 participants and the reduced sample of N = 165 participants. Results were the same between the two analytic samples, and we thus report findings for the full sample.
Of the 104 participants who reported chronic pain and had pain diagnoses in their VA medical record, 87 (84%) were diagnosed with chronic neck or joint pain, 75 (72%) with chronic low back pain, 71 (68%) with rheumatism/arthritis, 25 (24%) with migraine headaches, 11 (11%) with neuropathic pain, nine (9%) with fibromyalgia, and five (5%) with inflammatory bowel disease. Seventeen participants (16%) had a single chronic pain diagnosis, 25 (24%) had two pain diagnoses, and 62 (60%) had three or more pain diagnoses. On average, patients reported, being diagnosed with chronic pain for 13.9 years (SD = 10.8 years, Median = 11.0 years). Fifty-eight participants (56%) with chronic pain self-reported taking prescription opioid medications in the past 30 days.
Exploratory analyses indicated that nearly all of the 110 participants with chronic pain had tried over-the-counter (96%) and opioid (90%) pain medications in the past, while a lesser majority had tried non-pharmacological pain treatments such as physical therapy (79%), TENS (58%), massage (57%), chiropractic treatment (52%), and relaxation training/stress management (69%). Slightly less than half (43%) of participants had tried herbal medicines to manage chronic pain, while approximately one-third had tried acupuncture (36%) and group pain management classes (32%), respectively. The use of multiple pain management treatment strategies in the past was the norm rather than the exception. No participant reported using 0–1 treatment strategies in the past, 13 (12%) reported using 2–3 strategies, 49 (44%) reported using 4–6 strategies, and 48 (44%) reported using 7–10 strategies.
Tables 1 and and22 provide descriptive statistics for demographic, quality of life, psychiatric, substance use, medical comorbidity, disability, and liver disease severity variables for the full sample and the subsamples of participants with and without chronic pain. Participants with chronic pain were more likely than participants without chronic pain to be single, χ2(3, n = 170) = 15.98, p < 0.01, but the groups did not differ on any other demographic variable, prevalence of major depressive disorder or PTSD, or current alcohol or substance abuse/dependence. Participants with chronic pain reported lower physical quality of life than participants without chronic pain, t(169) = 4.87, p < 0.01, but the groups did not differ on mental health quality of life. Relative to participants without chronic pain, those with chronic pain reported higher pain severity, t(168) = 7.38, p < 0.01, greater pain interference, t(163) = 7.47, p < 0.01, experienced more illness due to medical conditions as assessed by the Charlson Comorbidity Index, t(169) = 2.63, p < 0.01, and were more likely to have a VA service connected disability, χ2 (1, n = 171) = 9.17, p < 0.01. Participants with chronic pain did not differ from those without chronic pain on liver disease severity as assessed by APRI.
After controlling for participant age, gender, marital status, Charlson Comorbidity Index, VA service connected disability status, APRI, and diagnoses of current major depressive disorder, PTSD, alcohol abuse/dependence, and substance abuse/dependence, the total number of inpatient stays in the past five years was greater for patients with chronic pain relative to those without chronic pain (OR = 2.58 [1.46, 4.56]), as was the length of hospital stays for psychiatric reasons (OR = 5.50 [3.37, 8.99]). In addition, participants with chronic pain had significantly more outpatient visits with Primary Care (OR = 1.73 [1.15, 2.59]), Physical Therapy (OR = 9.57 [4.79, 19.11]), and Occupational Therapy (OR = 2.72 [1.00, 7.48]) in the past five years than those without chronic pain. Participants without chronic pain had no visits to a specialty pain clinic in the past five years, thus precluding inferential comparisons with participants with chronic pain (Mean = 0.77, SD = 2.13) on this utilization outcome variable. Participants with and without chronic pain did not differ significantly on number of days hospitalized for general medical procedures, after controlling for covariates (OR = 1.12 [0.54, 2.32]). See Table 3 for descriptive statistics of medical service utilization.
To evaluate the possible confounding effect of length of pain diagnosis on medical service utilization in the past five years, we conducted parallel analyses after excluding 26 participants who reported being diagnosed with chronic pain in the past five years (i.e., relatively new diagnoses of chronic pain). Results did not differ from the full sample. Details of these analyses are available upon request from the first author.
This study examined chronic pain prevalence, health service utilization, and pain treatment experience among a sample of U.S. military veterans with HCV infection. Nearly two-thirds of participants reported having been diagnosed with chronic pain and had corroborating pain diagnoses in their medical records. This finding is in line with prior studies that have estimated prevalence of chronic pain among patients with HCV to range from 67%–83% (5–8). Among participants diagnosed with chronic pain, 84% had two or more diagnoses of pain documented in the medical record.
Compared to participants with HCV but no chronic pain, veterans with comorbid HCV and chronic pain reported lower physical quality of life and utilized more health services including number of inpatient hospital stays and number of days on inpatient psychiatric wards, even after adjusting for demographic variables, comorbid medical conditions, disability indicators, liver disease severity indicators, depression, PTSD, and substance use covariates. In addition, participants with chronic pain had more total outpatient visits with primary care providers, physical therapists, occupational therapists, and specialty pain providers. These findings indicate that pain is associated with increased health service utilization among patients with HCV. The extensive use of both inpatient and outpatient services by patients with comorbid HCV and chronic pain has significant cost implications, particularly considering that pronounced health service use disparities between patients with and without chronic pain were evidenced with inpatient hospital stays, which incur the greatest cost to a medical organization and the greater healthcare system.
Exploratory descriptive data on past pain treatment use indicated that patients with HCV and chronic pain had tried a significant number of pain treatments in the past, with over 90% reporting prior use of pharmacotherapy, including prescription opioid medications. Others have observed lower rates of pharmacotherapy use in the past 12 months among both non-veteran and veteran patients with chronic pain. For example, 33% of patients with chronic pain identified through a population survey were currently taking analgesic medication to manage pain (40). Among U.S. veterans with chronic pain, 41% were prescribed opioid medication by a VA provider over a 12-month prospective period (41), while others have found that 30% of U.S. veterans with comorbid HCV and chronic pain were prescribed opioid medication in the past 12 months (8). In the current study, we assessed reported lifetime use of pharmacotherapy, which likely accounts for the higher proportion of patients reporting pharmacotherapy for pain management.
Although 90% of participants reported past use of prescription opioids, fewer (56%) reported current use. The reasons for reduced usage of opioids among these patients remain elusive due to limitations in our data. It is possible that opioids may have been discontinued for a number of reasons, including medication side effects, limited analgesia, concurrent substance use, or some other factor. Providers may be reticent to prescribe opioid medications to patients with histories of alcohol or substance abuse (42–43), of which this patient sample was almost entirely comprised. Alternatively, some patients may choose not to take opioids due to concerns about potential addiction, despite providers’ willingness to prescribe these medications. Still other patients may have undergone trials of opioid medications in the past, but had little pain relief and/or experienced significant medication-related side effects (44).
Many non-pharmacological pain treatments had also been used by the majority of patients with chronic pain in the current study, albeit to a lesser extent than pharmacological pain treatments, perhaps due to disproportional referrals made by providers for pharmacological and non-pharmacological pain treatments (45), although data were not available in our study to assess rates of referral. The moderately high rates of non-pharmacological pain treatment use by patients with comorbid HCV and chronic pain is consistent with findings from non-veteran samples of patients with chronic pain, which indicate that the majority of patients report prior or current use of various complementary pain management treatments (e.g., chiropractic, massage, acupuncture; 46). This reported use of various non-pharmacological pain treatment strategies may be due to the accessibility and affordability of these treatments for veterans within the VA system such as physical therapy, TENS, and cognitive-behavioral therapy for chronic pain.
This study has several limitations. Namely, this study was conducted exclusively with U.S. military veterans treated at a single hospital within the VA healthcare system, and results may not generalize to other populations of patients with HCV. In addition, patients with HCV in this study did not have advanced liver disease or had not been diagnosed with hepatocellular carcinoma. We also excluded participants who had been treated with interferon for HCV. As such, it is unclear if findings on increased medical service utilization for patients with comorbid HCV and chronic pain would be replicable in a sample of patients with HCV that is more representative of the full HCV disease severity and treatment spectra.
We extracted data from a single VA hospital system and did not account for the possibility that participants may have received medical care at other VA or non-VA hospitals, and we are thus limited in our ability to interpret medical service utilization. In addition, we did not employ an HCV-negative control group to test whether HCV and chronic pain has a cumulative effect on health service utilization. Future studies could employ such four-group designs (i.e., HCV-negative without chronic pain, HCV-negative with chronic pain, HCV-positive without chronic pain, and HCV-positive with chronic pain), selecting random, rather than convenience, samples of patients with HCV and obtain health service data from multiple medical sources to evaluate this research question.
An additional limitation concerns the cross-sectional nature of this study. As such, the presence of pain cannot be said to cause increased use of health services. A prospective longitudinal study would be required to provide further evidence of this relationship. Furthermore, we did not evaluate the extent to which chronic pain contributed to a given inpatient or outpatient medical encounter, and future studies should help elucidate the relative contributions of pain to medical treatments received.
Finally, data on past pain treatment use were limited in several ways. Our data should be viewed as preliminary as we could not elucidate patterns of past chronic pain treatment use, acceptability of treatments used in the past, or reasons for discontinued use. Future studies should employ treatment use measures that can fill this existing gap in the literature.
This study has several important strengths. It is the first to examine the relationship between pain and health service utilization in a sample of patients diagnosed with HCV. The high rates of chronic pain among patients with HCV, coupled with increased health service utilization, indicate that engagement in multi-modal pain treatment may be a viable method of addressing patients’ pain and reducing the burden on the greater healthcare system. Future studies that examine the efficacy of pain treatment for patients with comorbid HCV and chronic pain appear warranted.
This material is the result of work supported with resources and the use of facilities at the Portland VA Medical Center. This study was supported in part by award K23DA023467 from the United States National Institute on Drug Abuse to Dr. Morasco. The authors appreciate the assistance of Jonathan Duckart, MPS, for extracting data from the electronic medical record. The content of this manuscript is solely the responsibility of the authors and does not necessarily represent the official views of the Department of Veterans Affairs or the National Institute on Drug Abuse.
Conflict of Interest: No author reports having any financial or other potential conflict of interest with this study.