Insight into chronic pain in the United States: descriptive results from the Prescription Opioid Misuse and Abuse Questionnaire (POMAQ) validation study

Abstract

Objective

A chronic pain patient sample living in the United States who participated in a cross-sectional study to evaluate the validity and reproducibility of the Prescription Opioid Misuse and Abuse Questionnaire is characterized.

Methods

Patients with chronic pain identified through electronic medical records as refilling at least one opioid prescription within the prior 3 months were recruited from five United States Department of Defense Military Health System clinics. Patients completed the Prescription Opioid Misuse and Abuse Questionnaire, Brief Pain Inventory-Short Form, Medical Outcomes Study: 36-item Short Form, and sociodemographic questions online. Clinical characteristics and electronic medical records for 1 year prior to consent were collected.

Results

809 (86.2%) participants completed the Prescription Opioid Misuse and Abuse Questionnaire. Mean (± standard deviation) age was 55.4 ± 12.7 years; the majority female (55.5%) and white (74.8%). Mean duration of chronic pain was 14.7 ± 10.5 years; the most common pain conditions were lower back pain (76.6%), neck or shoulder pain (60.3%), and osteoarthritis (38.7%). The most commonly prescribed opioids were oxycodone (35.7%), tramadol (34.5%), and hydrocodone (26.9%); 54.8% took one opioid, 44.9% took 2 or more opioids.

Discussion

Participants’ health status was poor; pain severity and interference were moderate. Electronic medical record data revealed high healthcare resource utilization. This chronic pain population was severely impacted by their pain condition(s).

Keywords:

• Prescription Opioid Misuse and Abuse Questionnaire (POMAQ)
• chronic pain
• prescription opioids
• prescription opioid misuse

Introduction

Estimated to impact anywhere from 25 million to 100 million Americans, chronic pain is a significant public health problem in the United States (US)^1,2. Chronic pain is a major contributor to national rates of morbidity, mortality, and disability and is one of the most common reasons for physician visits¹. Contributing an estimated $560 billion annually in the United States to direct medical costs, lost work productivity, and disability programs¹ and linked to lower levels of life satisfaction, reduced health-related quality of life, and higher rates of suicide³, the impacts of chronic pain extend beyond economics and underlie numerous psychological and social issues. The diverse consequences associated with chronic pain underscore the importance of appropriate and compassionate patient care.

Although there is little evidence to support long-term opioid use, the long-term use of prescription opioid pain medications in the treatment of moderate to severe chronic pain has become a therapeutic mainstay for select subgroups of patients⁴. Unfortunately, long-term use can result in problematic patterns of use and non-therapeutic use potentially leading to opioid use disorder (OUD)⁵. Chronic pain is an extraordinarily complex condition that is influenced by a multitude of factors, including biologic processes, and psychological, social, and environmental influences⁶. Adding to this complexity are the various patient-level risk factors that are associated with the development of OUD, including the history of substance abuse and mental health disorders⁷. Despite a decline in opioid prescriptions in recent years due to general concerns surrounding the national opioid crisis and increased awareness of OUD, opioid prescriptions remain very high in the United States⁸.

Increased prescribing of opioid medications in the United States over the past 20 years has been correlated with the development of a national crisis of increasing addiction to and misuse of both prescription and non-prescription opioids, resulting in the Department of Health and Human Services (HHS) declaring opioid use disorder a public health emergency in 2017⁹. In response to the urgent, complex, and intertwined public health priorities of effective chronic pain treatment and prevention of opioid misuse and abuse, the US Food and Drug Administration (FDA) has launched several initiatives to evaluate and improve research methods and measurement in the evaluation of potential new pain therapies¹. In the absence of a “gold standard” measure of prescription opioid misuse/abuse, the FDA requested a study as part of their post-marketing requirement (PMR) of pharmaceutical makers of extended-release/long-acting prescription opioids, to develop and validate a measure of the opioid-related adverse events of misuse and abuse among patients with chronic pain prescribed long-term opioid therapy^10,11 (see Supplemental Digital Content). While there are several widely used patient screeners for opioid risk^12–17, these instruments are not validated for use in a chronic pain population or are not designed to assess intentionality and/or do not distinguish between behaviors of misuse and abuse. To address the FDA’s PMR, the Prescription Opioid Misuse and Abuse Questionnaire (POMAQ) was derived from the Self-Reported Misuse, Abuse and Diversion of Prescription Opioids (SR-MAD)¹⁶, with extensive modifications and additional questions included to address the specific PMR research needs. The POMAQ was designed to fill the need for assessment of misuse/abuse and intentionality of prescription opioid use.

The POMAQ underwent content validation via qualitative interviews (Protocol 3033-3¹⁸), but quantitative evidence of its validity as a measure of misuse and abuse of prescription opioid medication had yet to be established. Therefore, a validation study (Protocol 3033-4), described herein, was implemented to evaluate the validity and reproducibility of the POMAQ to identify opioid abuse and misuse behaviors among patients who have chronic pain requiring long-term opioid use and to identify patterns of behaviors that commonly co-occur within these individuals. The purpose of this paper is to present the descriptive results of the POMAQ validation sample in terms of demographics and clinical characteristics, thereby depicting patients living with chronic pain in the United States and sharing insights gained from this population. Complete results from the POMAQ validation study, including evaluation of its construct validity and reproducibility, will be presented in separate publications.

Methods

Key definitions

The working definitions of misuse and abuse that provided the conceptual framework for the POMAQ were obtained from the Analgesic, Anaesthetic, and Addiction Clinical Trials, Translation, Innovations, Opportunities, and Networks (ACTTION) review¹⁹. These terms were defined as follows:

Misuse: Intentional use of a drug for therapeutic purpose (to reduce an aversive symptom or state), inappropriately outside label directions or in a way other than prescribed or directed by a healthcare practitioner; this includes patients using a drug for a condition different from that for which the drug is prescribed, patients taking more drugs than prescribed, or at different dosing intervals.

Abuse: Intentional use of a drug for non-therapeutic purposes, repeatedly or sporadically, for the purpose of achieving a positive psychological or physical effect.

Study design

The study population was recruited from the US Department of Defense (DoD) Military Health System (MHS) which serves approximately 10 million beneficiaries (i.e. active duty military, national guard members/reservists, family members, and military retirees) and is supported by a robust electronic medical record (EMR) system. Participants were recruited from five DoD clinics located across the US including: Naval Medical Center Portsmouth (Portsmouth, VA); Naval Medical Center San Diego (San Diego, CA); Wright-Patterson Medical Center (Dayton, OH); Walter Reed National Military Medical Center (Bethesda, MD); and San Antonio Military Medical Center (San Antonio, TX). Dorrance and colleagues found the DoD population to be representative of the insured US population in terms of age, gender, occurrence of common conditions and overall health²⁰. Various geographic locations were utilized to increase the generalizability of the sample. Patients were eligible from clinics within each of the selected centers provided they had a chronic pain diagnosis, were currently taking prescription opioids, and had refilled at least one prescription for opioid pain medication within the prior 3 months. Potentially eligible patients were identified by site staff through the DoD MHS EMR database and screened for study eligibility over the phone or in-person.

To be eligible, patients had to be ≥18 years, diagnosed with a chronic (≥3 months) pain condition requiring long-term treatment with opioids, willing to provide informed consent and a urine sample, and able to participate in an internet-based survey and telephone interview in English. Patients diagnosed with a terminal illness or who were currently active-duty service members of the military were not eligible to participate. Eligible and interested participants were scheduled for an initial clinic visit during which they provided written informed consent. Upon obtaining consent from the patient, the site coordinators assigned the participant a unique study identification (ID) and enrolled the participant into the web-based enrolment portal.

All recruitment procedures complied with the current Health Insurance Portability and Accountability Act (HIPAA) regulations. The protocol was approved by an Institutional Review Board (IRB) at each participating site, and a Certificate of Confidentiality (CoC) from the National Institutes of Health was obtained.

Procedures

Upon providing consent, participants were given a study packet that included a welcome letter, study information, and directions on how to complete a series of questionnaires via a secure website on their own time and at a location of their choice. The study packet included the participant’s unique study ID and registration number which allowed them access to the online survey and linked their responses to their de-identified EMRs. iPads were also available at clinical sites for participants to use to complete the survey onsite if desired. If completed at the site, patients were provided with a private location and unrestricted time to complete the survey. The survey battery included: the POMAQ, Brief Pain Inventory-Short Form (BPI-SF), Medical Outcomes Study: 36-item Short Form Health Survey Instrument (MOS SF-36), and sociodemographic questions. Following the clinic visit, participants were given up to 21 days to complete the series of questionnaires included in the survey. After the participant completed all the internet-based questionnaires, they received a $25 gift code via e-mail. Clinic staff also completed a clinical case report form (CRF) for each participant enrolled in the study.

Measures

Prescription Opioid Misuse and Abuse Questionnaire (POMAQ)

The POMAQ was designed as a series of questions with various response patterns about behaviors divided into 19 specific items. The questions specifically focused on behaviors related to misuse and abuse and the intention behind each behavior to identify prescription opioid misuse and abuse as well as prescription opioid diversion behaviors. A behavior or combination of behaviors were classified as opioid misuse or abuse (or both) based upon how the person responded to the intent of the specific behavior. Questions assessing behaviors over the previous one-year timeframe were used at baseline. Scoring guidelines for the POMAQ will be detailed in the algorithm validation manuscript (in development).

Medical Outcomes Study: 36-item Short Form Health Survey Instrument (MOS SF-36)

The MOS SF-36 is a self-administered, validated questionnaire that measures the following eight health aspects – physical functioning, role limitations due to physical problems, social functioning, bodily pain, mental health, role limitations due to emotional problems, vitality, and general health perception^21–24. Higher scores indicate better health-related quality of life (HRQL). The MOS SF-36 has been used extensively in numerous clinical trials and studies with published reliability and validity^22,25 and provides benchmark generic health status data on the patient sample.

Brief Pain Inventory – Short Form (BPI-SF)

The patient-completed BPI-SF assessed pain intensity through four questions (pain at its worst, least, average and now [current pain]); these four questions were analyzed separately and averaged to obtain a composite Pain Severity score. Further, the BPI-SF’s pain interference item with seven sub-questions assessed the degree that pain interferes with various daily activities²⁶. All items used an 11-point numerical rating scale, with higher scores indicating greater pain intensity or interference. The BPI-SF has been used, tested, and validated extensively in various patient populations^27,28 and provides benchmark data regarding the level of pain the participants experience. A 24 hour recall period was used for each item except for the questions on average and current pain intensity which did not stipulate a specific timeframe for the response.

Sociodemographic form

Participants completed a brief sociodemographic questionnaire after completing the other patient-reported outcome measures. The sociodemographic form collected the participant’s age, ethnicity, living situation, employment, and education.

Clinical form

Based on EMR data and clinical notes, clinical site staff completed a clinical CRF to document basic information related to opioid use history (e.g. type of opioid pain medication, duration of opioid use, known history of misuse or abuse) for each participant who completed a study visit.

Electronic medical record (EMR) data

Participants’ available EMR data were obtained from the DoD military healthcare data repository for one year prior to the consent date and included data on diagnoses, healthcare resource utilization, and prescriptions filled through this healthcare system.

Statistical analysis

All statistical analyses were performed using SAS version 9.4 (SAS Institute, Cary, NC). Descriptive statistics were derived for demographic and clinical variables to describe the study population. Descriptive statistics were derived for each POMAQ item as well as the other ancillary study measures including the MOS SF-36 and BPI-SF. Scores were calculated using each instrument’s scoring manual.

Results

Screening and sample disposition

A total of 3263 patients were screened; 1588 patients (48.7%) declined to participate due to lack of interest or time, or they disagreed with the study purpose; 604 (18.5%) were ineligible due to a variety of reasons (e.g. no longer taking opioids, too ill, living too far away) while 133 (4.1%) were eligible and willing to participate but were not enrolled due to difficulties with scheduling and/or site logistics; and 938 (28.7%) were determined to be eligible, consented to participate, and were enrolled. The POMAQ survey was completed by 809 (86.2%) of those who consented to be in the study. Of those who did not complete the POMAQ (13.8%), the majority (89.9%) were non-responsive to follow-up phone calls from the site, and the remainder (10.1%) withdrew from the study.

Self-reported demographic and clinical characteristics

The mean age ± standard deviation (SD) of participants was 55.4 ± 12.7 years (Table 1) with the majority being female (55.5%). Most participants were of non-Hispanic/Latino ethnicity (n = 733, 90.6%), and 74.8% (n = 605) of the sample were White/Caucasian. Participants tended to be employed full-time (n = 167, 20.6%), retired (n = 231, 28.6%), or on disability (n = 200, 24.7%). The majority either had some college education (n = 295, 36.5%) or a college degree (n = 273, 33.7%).

Table 1. Participant-reported demographic and clinical characteristics.

Characteristic	Total (N = 809)
Age, mean years ± SD	55.4 ± 12.7
Range (min–max)	(20.0–88.0)
Gender, n (% female)	449 (55.5%)
Ethnicity, n (%)
Hispanic or Latino	70 (8.7%)
Not Hispanic or Latino	733 (90.6%)
Missing	6 (0.7%)
Race, n (%)
White or Caucasian	605 (74.8%)
Black or African American	84 (10.4%)
Asian	9 (1.1%)
Native Hawaiian or other Pacific Islander	2 (0.2%)
American Indian or Alaska Native	10 (1.2%)
Other	93 (11.5%)
Missing	6 (0.7%)
Employment, n (%)
Employed, full-time	167 (20.6%)
Employed, part-time	58 (7.2%)
Homemaker	92 (11.4%)
Student	12 (1.5%)
Unemployed	35 (4.3%)
Retired	231 (28.6%)
Disabled	200 (24.7%)
Other	13 (1.6%)
Missing	1 (0.1%)
Education, n (%)
Elementary/primary school	5 (0.6%)
Secondary/high school	85 (10.5%)
Some college	295 (36.5%)
College degree	273 (33.7%)
Postgraduate degree	141 (17.4%)
Other	8 (1.0%)
Missing	2 (0.2%)
Comorbid conditions (n, %)^a
None	71 (8.8%)
Angina (chest pain)	44 (5.4%)
Arthritis	541 (66.9%)
Cancer	80 (9.9%)
COPD/emphysema	44 (5.4%)
Congestive heart failure/heart disease	22 (2.7%)
Diabetes	186 (23.0%)
Hypertension (high blood pressure)	372 (46.0%)
Kidney problems	92 (11.4%)
Vascular disease	46 (5.7%)
Other	348 (43.0%)
Missing	5 (0.6%)
Current pain condition, (n, %)^a
Bone break or fracture	87 (10.8%)
Central pain	111 (13.7%)
Fibromyalgia	172 (21.3%)
Low back pain	620 (76.6%)
Migraine or other chronic headache	256 (31.6%)
Neck or shoulder pain	488 (60.3%)
Osteoarthritis	313 (38.7%)
Pain related to cancer	35 (4.3%)
Painful diabetic neuropathy	78 (9.6%)
Phantom pain	29 (3.6%)
Post herpetic neuralgia	10 (1.2%)
Post-operative pain	138 (17.1%)
Rheumatoid arthritis	120 (14.8%)
Sickle cell pain	1 (0.1%)
Visceral pain	14 (1.7%)
Other neuropathic pain (nerve damage)	296 (36.6%)
Other	224 (27.7%)
Missing	2 (0.2%)

Abbreviations. COPD, chronic obstructive pulmonary disease; SD, standard deviation.

^aNot mutually exclusive.

Arthritis was the most commonly self-reported comorbid condition (n = 541, 66.9%) followed by hypertension (n = 372, 46.0%). Only 8.8% (n = 71) reported no comorbid conditions. Many participants (n = 348, 43.0%) reported “Other” comorbid conditions. The three most commonly self-reported pain conditions (not mutually exclusive) were lower back pain (n = 620, 76.6%), neck or shoulder pain (n = 488, 60.3%), and osteoarthritis (n = 313, 38.7%). Slightly over a quarter of patients (n = 224, 27.7%) reported “Other” pain conditions.

Site-reported clinical characteristics

Based on the staff-completed clinical CRF, participants had been seen at their respective medical practices for an average of 6.2 ± 7.8 years (range: 0.0–50.0 years, median: 3.0 years) (Table 2). The vast majority (n = 806, 99.6%) of participants were currently taking opioid pain medications, with a mean duration of opioid medication use of 7.3 ± 5.6 years (range: 0.2–47.0 years, median 7.0). The remaining three (0.4%) had previously taken opioids but were no longer taking opioids at the time they completed the questionnaire. The most commonly prescribed current opioid pain medications were oxycodone alone or in combination with another medication (n = 289, 35.7%), tramadol alone or in combination with another medication (n = 279, 34.5%), and hydrocodone alone or in combination with another medication (n = 218, 26.9%); 45% were currently prescribed two or more prescription opioids.

Table 2. Site-reported clinical characteristics.

Characteristic	Total (N = 809)
Duration at clinical practice, mean years (SD)	6.2 (7.8) 0.3–20.0
Range (min - max)	(0.0–50.0)
Median	3.0
Currently taking opioid pain medication, n (%)	806 (99.6%)
Duration of opioid medication, mean years (SD)	7.3 (5.6)
Range (min–max)	(0.2–47.0)
Median	7.0
Current opioid pain medication, n (%)^a
Hydrocodone alone or in combination with another medication	218 (26.9%)
Hydromorphone	40 (4.9%)
Oxycodone alone or in combination with another medication	289 (35.7%)
Methadone	23 (2.8%)
Codeine alone or in combination with another medication in tablet or solution	21 (2.6%)
Oxymorphone	17 (2.1%)
Morphine	68 (8.4%)
Fentanyl	47 (5.8%)
Buprenorphine	28 (3.5%)
Tramadol alone or in combination with another medication	279 (34.5%)
Tapentadol	26 (3.2%)
Meperidine (Demerol^®)	0 (0.0%)
Other prescription opioid pain medication	7 (0.9%)
None of the above	0 (0.0%)
No known misuse or abuse opioids or other substances	757 (93.6%)
Known misuse or abuse opioids or other substances
Known Opioid Abuse^a	19 (2.3%)
Known Abuse of other substances^a	28 (3.5%)
Duration of abuse of opioids or other substances, mean years (SD) (N = 16)	13.3 (14.6)
Range (min–max)	(0.4-51.0)
Known Opioid Misuse	11 (1.4%)
Known Misuse of other substances	5 (0.6%)
Duration of misuse of opioids or other substances, mean years (SD) (N = 11)	6.4 (6.5)
Range (min–max)	0.3-20.0

Abbreviations. SD, standard deviation.

^aNot mutually exclusive.

Based on their medical records, most participants (n = 757, 93.6%) had no reported prior history of misuse or abuse of opioids or other substances. However, for those with a history of abuse (n = 19 for abuse of opioids and n = 28 for abuse of other substances) or misuse (n = 11 for misuse of opioids and n = 5 for misuse of other substances), the mean duration of abuse or misuse was 13.3 ± 14.6 years and 6.4 ± 6.5 years, respectively.

Health status as reported on the MOS SF-36 and BPI-SF

MOS SF-36 domain scores ranged from 0 to 100 with higher scores indicating more favorable health status. Mean ± SD domain scores demonstrated lower scores on the physical health domains compared to the mental health domains, with scores ranging from 17.8 ± 30.9 (role limitations due to physical health) to 69.0 ± 20.6 (mental health) (Figure 1), which is significantly lower than the general population norm-based scores of 50 with a standard deviation of 10.

Figure 1. Descriptive Statistics for the MOS SF-36. ¹Scores have a possible range of 0–100. Higher scores indicate more favorable health status. The general population norm-based score is 50 with a standard deviation of 10. ²The scoring algorithm used to calculate the MOS SF-36 scores was developed by Hays et al. [41]. This version of the scoring algorithm does not incorporate weights for calculating the mental component summary (MCS) and physical component summary (PCS). The updated weights for the MCS and PCS are copyrighted and not publicly available for use. As such, in these analyses, MCS and PCS scores were not calculated for the MOS SF-36.

Scores for each item of the BPI-SF range from 0 to 10, with 10 indicating more pain and/or interference with activities. Mean ± SD scores for the four individual BPI-SF severity items ranged from 3.3 ± 2.0 (least pain over past 24 hours) to 6.9 ± 2.1 (worst pain over past 24 hours) while the Pain Severity score (average of the four items) was 4.9 (Figure 2).

Figure 2. Descriptive Statistics for the BPI-SF. ¹BPI-SF item scores range from 0 to 10, with higher scores indicating greater pain severity or interference. Pain Severity is a composite of the four pain items (a mean severity score). Pain interference is scored as the mean of the seven interference items if more than 50%, or four of seven, of the total items have been completed on a given administration

Mean ± SD scores on the BPI-SF interference items ranged from 3.8 ± 3.1 (relations with others) to 5.9 ± 3.0 (normal work). The mean pain interference summary score (a mean of the seven interference items if >50% of items were completed) was 5.3 ± 2.5.

POMAQ item characteristics

More than half the sample (54.8%) reported currently taking only one prescription opioid pain medication on item 1 of the POMAQ, while 44.9% reported taking two or more opioids. Key findings for POMAQ behaviors reported as having occurred in the past year (prior to the date of POMAQ completion) yielded a range of prevalence from 0% (purchasing or stealing a prescription pad) to 43.6% (taking less opioids than prescribed) with most POMAQ behaviors being reported by 2–20% of the sample as listed in Table 3. Most participants (n = 671, 82.7%) reported not taking any street drugs in the past year, and of the 135 participants who did, marijuana (n = 81, 10.0%) and sleeping pills (n = 40, 4.9%) were the drugs most widely reported.

Table 3. POMAQ results (items 1–8 and 10–19).

Question	Total (N = 809)
Q1: Opioids prescribed in past 3 Months (n [%])
Hydrocodone alone or in combination	247 (30.5%)
Hydromorphone	65 (8.0%)
Oxycodone alone or in combination	316 (39.1%)
Methadone	28 (3.5%)
Codeine alone or in combination	48 (5.9%)
Oxymorphone	17 (2.1%)
Morphine	97 (12.0%)
Fentanyl	58 (7.2%)
Buprenorphine	37 (4.6%)
Tramadol alone or in combination	293 (36.2%)
Tapentadol	35 (4.3%)
Other opioids	5 (0.6%)
Other prescription opioid pain medication	39 (4.8%)
None of the above	3 (0.4%)
Number of prescribed opioids (n [%])
0	3 (0.4%)
1	443 (54.8%)
2	280 (34.6%)
3 or more	83 (10.3%)
Q2: Have taken non-prescribed opioids	20 (2.5%)
Q3: Took less opioids than prescribed in past year	353 (43.6%)
Q4: Have you taken more than prescribed opioids	150 (18.5%)
Q5: Have changed or tampered with opioids	16 (2.0%)
Q6: Drank alcohol while taking opioid	233 (28.8%)
Q7a: Stimulant medications	50 (6.2%)
Q7b: Anti-anxiety medications	245 (30.4%)
Q7c: Sleeping pills	204 (25.3%)
Q7d: Antihistamines	171 (21.2%)
Q7e: Barbiturates	5 (0.6%)
Q7f: Antipsychotics or neuroleptics	51 (6.3%)
Q7g: Other sedatives	70 (8.7%)
Q8: Got opioids from someone other than a doctor or healthcare provider in past year	19 (2.3%)
Q9: Number of street drugs taken in past year
0	671 (82.9%)
1	100 (12.3%)
2	27 (3.3%)
3 or more	8 (1.0%)
Missing	3 (0.4%)
Q10: Did you visit more than 1 doctor or healthcare provider for more opioids in past year	86 (10.6%)
Q11: Have you visited more than 1 pharmacy to obtain opioid prescription	206 (25.5%)
Q12: Have you lost opioid or prescription	19 (2.3%)
Q13: Have your opioid or prescription been stolen	16 (2.0%)
Q14: Have you requested refills for prescribed opioids earlier in past year	107 (13.2%)
Q15: Have you visited emergency room or urgent care clinic for more opioids	29 (3.6%)
Q16: Have you used or tried to use a fake or changed prescription for opioids in past year	4 (0.5%)
Q17: Have you purchased or stolen a prescription or prescription pad or part of a prescription pad in past year	0 (0.0%)
Q18: Have you shared, sold or traded your opioids	17 (2.1%)
Q19: Do you think you have a problem with your prescription opioid medication (n, %)
Yes	38 (4.7%)
No	737 (91.1%)
I am not sure	34 (4.2%)

EMR results

Extracted EMR data demonstrated high healthcare resource utilization among this chronic pain patient population. Emergency room visits were reported for 40.8% of participants, while 14.3% had been hospitalized at least once in the year prior to the study (Table 4). In the year prior to study enrolment, the mean (± SD) number of visits per to a DoD Military Health System (MHS) outpatient facility was 33.4 (± 29.3) visits per participant.

Table 4. EMR data (for year prior to study start).

	Total (N = 809)
Average number of diagnoses per patient
Mean (SD)	15.9 (6.7)
Median (IQR)	15 (11–43)
Healthcare resource utilization
Number of outpatient visits per patient, mean (SD)	33.4 (29.3)
ER visits, n (%)	330 (40.8%)
Hospitalizations, n (%)	116 (14.3%)
Prescription utilization
Average number of opioid prescriptions filled, mean (SD)	15.4 (8.2)
Number of participants receiving other prescriptions, n (%)	807 (99.8%)

Abbreviations. ER, emergency room; IQR, interquartile range; SD, standard deviation.

The EMR data revealed a high number of unique International Classification of Diseases, Ninth Revision (ICD-9) diagnoses per patient with an average of 15.9 ± 6.7 diagnoses per person (diagnoses recorded in the 1 year prior to study), indicating the presence of many comorbid conditions. Examples of common comorbid conditions were anxiety (24.0%), major depression (28.8%), orthopaedic conditions (61.9%), neurologic conditions (58.3%), and sleep disorders (38.6%) (see Supplemental Digital Content Table S1). Nearly all participants (99.8%) were prescribed medications other than opioids. The most common of those medications included: gastrointestinal medication (77.1%), non-steroidal anti-inflammatory drugs (NSAIDs) (70.1%), cardiac medication (65.6%), anti-infectives (64.8%), antidepressants (62.2%), anticonvulsants (57.1%), and muscle relaxants (48.3%) (see Supplemental Digital Content Table S2). The mean number of opioid prescriptions filled per person over the past year was 15.4 ± 8.2.

Discussion

This paper describes the characteristics of a large sample (n = 809) of patients with chronic pain currently prescribed long-term treatment with opioids recruited through DoD MHS clinics. These patients were recruited to participate in a cross-sectional study (to be presented in separate publications) intended to evaluate the validity and reproducibility of the POMAQ as a measure of opioid-related adverse events of misuse and abuse among patients with chronic pain prescribed long-term opioid therapy. The self-reported general health status of patients in this sample was poor, as assessed by the MOS SF-36, while their pain severity and interference were moderate, based on the BPI-SF. The majority had lower back pain and/or neck or shoulder pain and nearly all (>90%) reported at least one comorbid condition. EMR data revealed numerous comorbid conditions, including a high prevalence of major depression (28.8%) and anxiety (24.0%), as well as substantial healthcare resource utilization and other prescription medication usage. In short, results reveal a chronically ill patient population suffering from numerous comorbid conditions layered on top of moderate to severe chronic pain. The complex health profile of this patient population requires extensive coordinated care through various avenues of the healthcare system.

The EMR data surrounding comorbidities and healthcare resource utilization in this study complement previously published work conducted among chronic pain populations^29–32. For example, Engel and colleagues examined a group of primary care patients with chronic back pain and found that having chronic pain along with other conditions, including multiple types of chronic pain, was linked to heightened use of the healthcare system and increased costs²⁹. A large retrospective analysis which examined insurance claims data of patients with chronic low back pain alongside matched controls noted that patients with chronic low back pain were characterized by greater comorbidity and economic burdens, including more than double the estimated direct medical costs compared to those without low back pain³⁰.

The high prevalence of depression and anxiety in this sample was not surprising given that painful and depressive symptoms often co-occur³³. Depression rates in this sample were rather high, similar to those found in previous studies evaluating depression and anxiety in chronic pain populations^30,34. Coyne and colleagues noted in a cross-sectional study among adult patients recruited through a primary care setting, levels of anxiety, depression and impact on patients’ usual activities, sleep, and mood were all higher among patients with chronic pain compared to those without chronic pain³⁴. Gore and colleagues also found higher rates of depression, anxiety, and sleep disorders in patients with chronic low back pain compared to matched controls without chronic low back pain³⁰. Importantly, despite the multiple comorbid conditions experienced by this study sample, a relatively small proportion of patients endorsed POMAQ items that would indicate either abuse or misuse behaviors surrounding their prescription opioid medication.

Despite chronic pain being one of the most common reasons for seeking medical care in the US¹, there is a lack of systematic screening, inconsistent recording of pain, and fractured delivery of patient care^35–37. As supported by findings within this study population and others, patients often pursue numerous primary and specialty care services to address their pain and associated conditions^1,38. This is evidenced by the high number of outpatient visits (33.4 ± 29.3) per participant over the one-year study period found in our study. This approach to pain management can lead to unfavorable outcomes and significantly increased healthcare resource utilization costs^38,39. Thus, it is imperative that the needs of chronic pain patients are met through careful, individualized and coordinated multidisciplinary care.

Several limitations of this study should be noted. First, this study population was recruited from five DoD military hospitals and their supporting clinics. Although noted by Dorrance and colleagues to be similar to the overall US population in terms of key demographics and common medical conditions, this sample is not representative of uninsured populations, which face a different set of health problems and utilize health care resources differently²⁰. It should also be noted that recruitment for this study was difficult with nearly half of the screened patients declining to enroll. This poor response rate can be attributed to a number of typical reasons, such as lack of time or interest and inconvenience as well as a lack of trust and/or a concern for losing access to their medication(s). There is no way to determine if the large number of patients who declined to participate in the study (48.7%) would have responded similarly or differently on the POMAQ or other measures compared to those who enrolled. Despite these limitations, the present sample was carefully and systematically recruited to provide an appropriate examination of the suitability of the POMAQ items for the population of chronic pain outpatients in the United States.

The description of this study sample revealed some important observations of chronic pain patients. Amid the growing focus on reducing exposure to opioids and enhancing rationalized prescribing patterns by the FDA and Centers for Disease Control and Prevention (CDC), patients’ fear of losing access to their necessary medication is top of mind for many of those battling chronic pain⁴⁰. While not the primary purpose of this study, our data present a useful picture of the complexity of pain care in this population. What we cannot learn from our data is the degree to which the co-morbidities and high level of healthcare utilization are due to the illnesses that create the chronic pain versus any possible issues that might arise from the patients’ chronic use of opioids. This is an important issue that clearly requires additional study. Regardless, monitoring for misuse/abuse behaviors is critical in evaluating whether a patient takes their prescription opioid medication properly; administering the POMAQ could assist physicians in identifying problematic behaviors and facilitating communication between patients and healthcare providers as potential issues arise. What is clear is that the association of any two findings in a cohort study does not demonstrate causality, and great care must be taken in the interpretation of this type of data, including in this study. Importantly, to effectively address the complex health profiles of chronic pain patients, their care will need to be delivered through a thoughtful and highly coordinated, multidisciplinary approach.

Conclusion

The large chronic pain population recruited for this cross-sectional study was severely impacted by their pain condition(s) but reported relatively few behaviors indicative of prescription opioid misuse or abuse.

Transparency

Declaration of funding

This project was conducted as part of a Food and Drug Administration (FDA)-required post-marketing study for extended-release and long-acting opioid analgesics and was funded by the Opioid Postmarketing Consortium (OPC) consisting of the following companies: Allergan; Assertio Therapeutics, Inc.; BioDelivery Sciences, Inc.; Collegium Pharmaceutical, Inc.; Daiichi Sankyo, Inc.; Egalet Corporation; Endo Pharmaceuticals, Inc.; Hikma Pharmaceuticals USA Inc.; Janssen Pharmaceuticals, Inc.; Mallinckrodt Inc.; Pernix Therapeutics Holdings, Inc.; Pfizer, Inc.; Purdue Pharma, LP.

Declaration of financial/other relationships

K. S. Coyne and B. M. Currie are employees of Evidera, a company that received funding from the OPC to conduct this study. At the time of the study, A. I. Barsdorf was employed by Pfizer, a member company of the OPC. At the time of the study, J. L. Poon was an employee of Evidera and is now an employee of Eli Lilly and Company. At the time of the study, J. Y. Mazière was an employee of PPD, a parent company to Evidera and is now an employee of Astellas. R. F. Pierson is an employee of Janssen, a member company of the OPC. S. H. Schnoll is an employee of Pinney Associates, a company that received funding from the OPC for time spent consulting on this study. S. F. Butler received funding from the OPC for time spent consulting on this study. J. T. Farrar is an employee of the University of Pennsylvania and received funding from the OPC for time spent consulting on this study to support his salary. H. J. Fisher is an employee of Health ResearchTx, a company that received funding from the OPC to conduct this study. All aspects of the study design, interpretation, and decision to submit for publication were determined by the authors.

Source of support

Research data derived from approved Naval Medical Research Unit-Dayton, Dayton, Ohio and Naval Medical Center, Portsmouth, Virginia IRB, protocol; number NAMRUD.2015.0004. The views expressed in this article are those of the author(s) and do not necessarily reflect the official policy or position of the Department of the Navy, Department of Defense, or the United States Government. CDR Michael Franks Jr. is a member of the U.S. Military. This work was prepared as part of his official duties. Title 17U.S.C. 105 provides that "Copyright protection under this title is not available for any work of the United States Government." Title 17U.S.C. 101 defines a United States Government work as a work prepared by a military service member or employee of the United States Government as part of that person's official duties.

Acknowledgements

The authors thank Kawthar Nakayima, Amara Tiebout, and Michael Grossi for their production and editorial assistance.