Pain Medicine 2015; 16: 112–118 Wiley Periodicals, Inc.
OPIOIDS, SUBSTANCE ABUSE & ADDICTIONS SECTION Original Research Articles Sex Differences in Chronic Pain Management Practices for Patients Receiving Opioids from the Veterans Health Administration
Elizabeth M. Oliva, PhD,* Amanda M. Midboe PhD,* Eleanor T. Lewis, PhD,* Patricia T. Henderson, MS,* Aaron L. Dalton MA, MSW,* Jinwoo J. Im, PhD,¶ Karen Seal MD, MPH,†,‡ Meenah C. Paik, MPH,* and Jodie A. Trafton, PhD*,§ *Department of Veterans Affairs, Center for Innovation to Implementation, Menlo Park, California; ‡ Integrated Care Clinic and Medical Practice Pain Clinic, San Francisco Veterans Affairs Medical Center, San Francisco, California; §Departments of Medicine and Psychiatry, University of California at San Francisco, San Francisco, California, ¶ Department of Psychiatry and Behavioral Sciences, Stanford University School of Medicine, Stanford, California, USA; †Industry & Strategy Department, Samsung Economic Research Institute, Seoul, South Korea Reprint requests to: Elizabeth M. Oliva, PhD, Health Science Specialist, VA Center for Innovation to Implementation, 795 Willow Rd. (152), Menlo Park, CA 94025, USA. Tel: 650-493-5000 x 2-23170 or 650-483-7515; Fax: 650617-2736; E-mail: [email protected]. Disclosure: The authors have no conflict of interest to report. The views expressed in this article are those of the authors and do not necessarily reflect the position or policy of the Department of Veterans Affairs or the United States government. Abstract Background. Women experience chronic pain and use pain-related health care at higher rates than men. It is not known whether the pain-related health
112
care female veterans receive is consistent with clinical practice guideline recommendations or whether receipt of this care differs between men and women. Objective. The aim of this study was to identify whether sex differences in chronic pain management care exist for patients served by the Veterans Health Administration (VHA). Design. Data on patient demographics, diagnostic criteria, and health care utilization were extracted from VHA administrative databases for fiscal year 2010 (FY10). Patients. Patients in this study included all VHA patients (excluding metastatic cancer patients) who received more than 90 days of a short-acting opioid medication or a long-acting opioid medication prescription in FY10 study. Measures. Multilevel logistic regressions were conducted to identify sex differences in receipt of guideline-recommended chronic pain management. Results. A total of 480,809 patients met inclusion criteria. Female patients were more likely to receive most measures of guideline-recommended care for chronic pain including mental health assessments, psychotherapy, rehabilitation therapy, and pharmacy reconciliation. However, women were more likely to receive concurrent sedative prescriptions, which is inconsistent with guideline recommendations. Most of the observed sex differences persisted after controlling for key demographic and diagnostic differences. Conclusions. Findings suggest that female VHA patients are more likely to receive an array of pain management practices than male patients, including both contraindicated and recommended
Sex Differences in Pain Management Practices polypharmacy. Quality improvement efforts to address underutilization of mental health and rehabilitative services for pain by male patients and polypharmacy in female patients should be considered. Key Words. Pain Veterans
Management;
Opioids;
Sex;
Introduction Studies of civilian populations in Western and nonWestern societies have revealed that the prevalence of chronic pain is higher in women than men [1–3]. Some evidence also indicates that there is greater pain-related disability in women compared with men [4,5]. Although a similar pattern has begun to emerge in research on pain in female veterans—e.g., a study of female veterans receiving care at a women’s health clinic at a northeastern Department of Veterans Affairs (VA) facility revealed a high proportion (78%) of women reporting persistent pain problems [6]—a recent series of studies suggests that considerations of service era and medical conditions may be necessary. Specifically, Haskell and colleagues in two separate studies found that despite female Operation Enduring Freedom (OEF)/Operation Iraqi Freedom (OIF) veterans who were within 1 year of their last deployment being less likely than male veterans to report chronic pain [7], the rate of female compared with male OEF/OIF veterans experiencing back pain, musculoskeletal conditions, and joint disorders increased each year up to 7 years post-deployment [8]. Given these sex differences in the experience of pain, it is not surprising that research on civilians has found sex differences in chronic pain management practices and use of health care. A civilian Dutch population-based study by Wijnhoven et al. [9] found that women with musculoskeletal pain used more health care and received more prescriptions for medication than men with musculoskeletal pain. Similarly, a study of two large health plans found some evidence that risky coprescribing of sedative hypnotics in conjunction with opioids is higher in women than men [10]. This same study also found women were more likely than men to receive opioid therapy, with older women being particularly likely to receive long-term opioid therapy [10]. Long-term opioid use may be more problematic for women than men given that women are more likely to take higher doses of opioids and may be at greater risk for polypharmacy and concomitant drug interactions [11]. Moreover, women are more likely to have risk factors such as psychiatric comorbidities that may increase their likelihood of opioid abuse and dependence [12]. In response to the rapidly increasing use of opioids to treat chronic pain, in 2010, the Department of Defense and VA jointly issued a Clinical Practice Guideline for the Management of Opioid Therapy for Chronic Pain with recommendations for treating chronic pain (hereafter
referred to as guideline-recommended practices) [13]. Although these guideline-recommended practices make no sex-specific treatment recommendations, they do outline a variety of pain management practices associated with safe and effective use of opioid medications. These practices include proper monitoring of patients via practices such as regular follow-up and urine drug tests, and avoiding sole reliance on opioids for pain management. This article will examine whether male and female patients who received chronic, short-acting or longacting opioid medications from the Veterans Health Administration (VHA) differed in receipt of chronic pain management guideline-recommended practices. Where sex differences are found, analyses will be conducted to identify whether these sex differences persist after accounting for key demographic and diagnostic factors. Methods Participants Female and male patients who received care at any VHA facility were included in this analysis if they had received more than a 90-day supply of any short-acting opioid medication(s) within the 12-month fiscal year 2010 (FY10) or any prescription for a long-acting opioid medication in FY10. The 90 days of supply criterion to define chronic, short-acting opioid use was established in consultation with an expert panel that included authors of the guideline [14]. Patients were excluded if they had a diagnosis of metastatic cancer. Short-acting opioids were defined as butorphanol, codeine, dihydrocodeine, hydrocodone, hydromorphone, meperidine, morphine, opium/belladonna, oxycodone, oxymorphone, pentazocine, and tapentadol. Long-acting opioids were defined as fentanyl, hydromorphone, methadone, sustained release morphine, sustained release oxycodone, sustained release oxymorphone, and sustained release tramadol. Data were obtained from VHA national administrative databases, including the VHA National Patient Care Database (NPCD), as well as VHA Decision Support System (DSS) pharmacy and laboratory files that were linked via scrambled social security numbers. This study was approved and overseen by the Stanford University Human Research Protection Program and the VA Palo Alto Health Care System Research and Development Committee. Metric Development The dependent variables in this project were developed to measure adherence across VHA facilities to recommendations in the 2010 Clinical Practice Guideline for the Management of Opioid Therapy for Chronic Pain (i.e., guideline-recommended practices; detailed descriptions and definitions of the variables are available in [14]). A panel of experts in opioid therapy and VHA pain management practice and policy met regularly for 1 year and guided the iterative development of individual metrics and metric definitions. These individuals are 113
Oliva et al. leading experts in the VHA on pain management, helped author the Clinical Practice Guideline, and had diverse professional and clinical backgrounds (e.g., pain medicine, pharmacy, behavioral medicine). A team of programmers with extensive experience with VHA administrative databases worked with the expert team throughout the development of the metrics to generate preliminary data and refine metrics as needed. These metrics were defined based on the following data elements: 1) dates of clinical encounters; 2) diagnoses treated (per International Classification of Diseases, 9th Revision, Clinical Modification [ICD-9] codes associated with encounters); 3) prescribed medications including fill date; 4) days’ supply; 5) formulation and daily dose; 6) labs ordered; 7) procedures conducted (per Current Procedural Terminology [CPT] and Healthcare Common Procedure Coding System [HCPCS] codes associated with encounters); and 8) location of care. Measures Demographic variables in this project were sex, age, marital status, and OEF/OIF status as coded in the FY10 VHA NPCD. Age was classified into three groups: 30 years old or younger, between 31 and 55 years old, and greater than 55 years old. Disorders were characterized based on ICD-9 diagnosis codes associated with VHA outpatient encounters or inpatient or residential stays as recorded in the NPCD. Any patient with at least one VHA health care visit for an ICD-9 diagnosis code in FY10 was characterized as positive for that disorder as follows: posttraumatic stress disorder (PTSD): 309.81; mood disorder (major depressive disorder or generalized anxiety disorder): 296.2–296.4 and 300.02; serious mental illness (SMI): 295, 297–299, 296.0–296.4, and 296.4 to 296.9 (affective psychosis, bipolar disorder, other psychosis, schizophrenia); and substance use disorder (SUD): 291–292.99, 303–305.1, 305.2–305.99 (alcohol use disorder, as well as the drug use disorders for opioids, cocaine, cannabis, amphetamines, barbiturates, and nicotine). Elixhauser comorbidity index categories of chronic pulmonary disease, cardiac arrhythmia, and liver disease (see Elixhauser et al. [15] for a listing of ICD-9 codes) were combined to create an index of medical frailty. These diagnoses were chosen specifically because they would be expected to influence opioid prescribing decisions. Metrics to Measure Adherence to GuidelineRecommended Practices A subset of the metrics described above was calculated for this study using FY10 data from the VHA NPCD and the VHA DSS dataset. The metrics measuring adherence to guideline-recommended practices included here are: 1. Ordering laboratory drug screens for (a) illicit drugs (e.g., cocaine, marijuana) and other illegal substances plus morphine-related opioid compounds (e.g., heroin), or (b) non-morphine opioids (e.g. hydrocodone, fentanyl). 2. Avoidance of sole reliance on opioids, as defined by receipt of (a) mental health assessment; (b) psycho114
therapy, including inpatient/outpatient and individual/ group treatment; (c) rehabilitation therapies including physical therapy, physical medicine and rehabilitation, recreational therapy, occupational therapy, pain clinic, or a prescribed exercise program; non-opioid pharmacotherapies including (d) anti-inflammatories, (e) tricyclic antidepressants (TCAs), (f) serotonin–norepinephrine reuptake inhibitors (SNRIs), and (g) anticonvulsants. 3. Side-effect management defined by prescription of a bowel regimen (i.e., a prescribed laxative or stool softener). 4. Pharmacy reconciliation defined by a pharmacy reconciliation encounter. 5. Contraindicated prescribing practices, as defined by receipt of (a) overlapping prescriptions for opioids and sedative medications (e.g., benzodiazepines, barbiturates, and carisoprodol [Soma]) and/or (b) single or overlapping prescriptions totaling more than 4 g per day of acetaminophen. Data Analysis Multilevel logistic regression was used to examine the two research questions with the primary independent variable being sex. Patients were linked to the VHA treatment facility where they received the majority of encounters or stays in FY10. The first set of analyses examined whether there were sex differences in receipt of the guidelinerecommended practices described above. Separate regressions were conducted with each of the guidelinerecommended practices included as dependent variables. Facility level rate of prescribing chronic, short-acting or long-acting opioids was included as a covariate and was defined as the percentage of total patients treated at the VHA facility who received a chronic, short-acting or longacting opioid prescription. The second set of analyses examined dependent variables that were identified as significant in the first set of analyses and, with sex as the independent variable, included the following covariates: 1) facility level rate of prescribing chronic, short-acting or long-acting opioids; 2) age category; 3) current marital status; 4) OEF/OIF status; 5) PTSD diagnosis; 6) SUD diagnosis; 7) mood disorder diagnosis; 8) SMI diagnosis; and 9) medical frailty. For any analyses with “yes” and “no” as the response, the reference group is “no.” Results Patient Characteristics In FY10, there were 480,809 VHA patients without metastatic cancer who received chronic, short-acting or long-acting prescription opioids. Of these patients, 389,335 (81%) received chronic, short-acting prescription opioids and 91,474 (19%) received long-acting prescription opioids. Table 1 describes the demographic characteristics of the sample. Overall, the sample was mostly male (93%), over age 55 (74%), married (53%), and diagnosed with a medical frailty (56%). Female VHA
Sex Differences in Pain Management Practices patients were significantly younger, more likely to be OEF/OIF veterans, less likely to be married, and had greater rates of PTSD, SMI, and mood disorder diagnoses, but lower rates of SUD diagnoses and medical frailty than male VHA patients (see Table 1).
Table 1
Primary Analyses The first set of multilevel logistic regression analyses revealed no significant sex differences between male and female VHA patients in prescription of a bowel regimen or in the use of urine drug screening for either illicit drugs or prescription opioids. For the remaining
Sample demographics by sex*
N Demographics Age >55 reference 31–55 30 OEF/OIF Currently married Yes Missing Diagnostic variables Medical frailty Posttraumatic stress disorder Substance use disorder Mood disorder Serious mental illness
Sample
Males
Females
480,809
449,006
31,803
P value
OPIOIDS, SUBSTANCE ABUSE & ADDICTIONS SECTION Original Research Articles Sex Differences in Chronic Pain Management Practices for Patients Receiving Opioids from the Veterans Health Administration
Elizabeth M. Oliva, PhD,* Amanda M. Midboe PhD,* Eleanor T. Lewis, PhD,* Patricia T. Henderson, MS,* Aaron L. Dalton MA, MSW,* Jinwoo J. Im, PhD,¶ Karen Seal MD, MPH,†,‡ Meenah C. Paik, MPH,* and Jodie A. Trafton, PhD*,§ *Department of Veterans Affairs, Center for Innovation to Implementation, Menlo Park, California; ‡ Integrated Care Clinic and Medical Practice Pain Clinic, San Francisco Veterans Affairs Medical Center, San Francisco, California; §Departments of Medicine and Psychiatry, University of California at San Francisco, San Francisco, California, ¶ Department of Psychiatry and Behavioral Sciences, Stanford University School of Medicine, Stanford, California, USA; †Industry & Strategy Department, Samsung Economic Research Institute, Seoul, South Korea Reprint requests to: Elizabeth M. Oliva, PhD, Health Science Specialist, VA Center for Innovation to Implementation, 795 Willow Rd. (152), Menlo Park, CA 94025, USA. Tel: 650-493-5000 x 2-23170 or 650-483-7515; Fax: 650617-2736; E-mail: [email protected]. Disclosure: The authors have no conflict of interest to report. The views expressed in this article are those of the authors and do not necessarily reflect the position or policy of the Department of Veterans Affairs or the United States government. Abstract Background. Women experience chronic pain and use pain-related health care at higher rates than men. It is not known whether the pain-related health
112
care female veterans receive is consistent with clinical practice guideline recommendations or whether receipt of this care differs between men and women. Objective. The aim of this study was to identify whether sex differences in chronic pain management care exist for patients served by the Veterans Health Administration (VHA). Design. Data on patient demographics, diagnostic criteria, and health care utilization were extracted from VHA administrative databases for fiscal year 2010 (FY10). Patients. Patients in this study included all VHA patients (excluding metastatic cancer patients) who received more than 90 days of a short-acting opioid medication or a long-acting opioid medication prescription in FY10 study. Measures. Multilevel logistic regressions were conducted to identify sex differences in receipt of guideline-recommended chronic pain management. Results. A total of 480,809 patients met inclusion criteria. Female patients were more likely to receive most measures of guideline-recommended care for chronic pain including mental health assessments, psychotherapy, rehabilitation therapy, and pharmacy reconciliation. However, women were more likely to receive concurrent sedative prescriptions, which is inconsistent with guideline recommendations. Most of the observed sex differences persisted after controlling for key demographic and diagnostic differences. Conclusions. Findings suggest that female VHA patients are more likely to receive an array of pain management practices than male patients, including both contraindicated and recommended
Sex Differences in Pain Management Practices polypharmacy. Quality improvement efforts to address underutilization of mental health and rehabilitative services for pain by male patients and polypharmacy in female patients should be considered. Key Words. Pain Veterans
Management;
Opioids;
Sex;
Introduction Studies of civilian populations in Western and nonWestern societies have revealed that the prevalence of chronic pain is higher in women than men [1–3]. Some evidence also indicates that there is greater pain-related disability in women compared with men [4,5]. Although a similar pattern has begun to emerge in research on pain in female veterans—e.g., a study of female veterans receiving care at a women’s health clinic at a northeastern Department of Veterans Affairs (VA) facility revealed a high proportion (78%) of women reporting persistent pain problems [6]—a recent series of studies suggests that considerations of service era and medical conditions may be necessary. Specifically, Haskell and colleagues in two separate studies found that despite female Operation Enduring Freedom (OEF)/Operation Iraqi Freedom (OIF) veterans who were within 1 year of their last deployment being less likely than male veterans to report chronic pain [7], the rate of female compared with male OEF/OIF veterans experiencing back pain, musculoskeletal conditions, and joint disorders increased each year up to 7 years post-deployment [8]. Given these sex differences in the experience of pain, it is not surprising that research on civilians has found sex differences in chronic pain management practices and use of health care. A civilian Dutch population-based study by Wijnhoven et al. [9] found that women with musculoskeletal pain used more health care and received more prescriptions for medication than men with musculoskeletal pain. Similarly, a study of two large health plans found some evidence that risky coprescribing of sedative hypnotics in conjunction with opioids is higher in women than men [10]. This same study also found women were more likely than men to receive opioid therapy, with older women being particularly likely to receive long-term opioid therapy [10]. Long-term opioid use may be more problematic for women than men given that women are more likely to take higher doses of opioids and may be at greater risk for polypharmacy and concomitant drug interactions [11]. Moreover, women are more likely to have risk factors such as psychiatric comorbidities that may increase their likelihood of opioid abuse and dependence [12]. In response to the rapidly increasing use of opioids to treat chronic pain, in 2010, the Department of Defense and VA jointly issued a Clinical Practice Guideline for the Management of Opioid Therapy for Chronic Pain with recommendations for treating chronic pain (hereafter
referred to as guideline-recommended practices) [13]. Although these guideline-recommended practices make no sex-specific treatment recommendations, they do outline a variety of pain management practices associated with safe and effective use of opioid medications. These practices include proper monitoring of patients via practices such as regular follow-up and urine drug tests, and avoiding sole reliance on opioids for pain management. This article will examine whether male and female patients who received chronic, short-acting or longacting opioid medications from the Veterans Health Administration (VHA) differed in receipt of chronic pain management guideline-recommended practices. Where sex differences are found, analyses will be conducted to identify whether these sex differences persist after accounting for key demographic and diagnostic factors. Methods Participants Female and male patients who received care at any VHA facility were included in this analysis if they had received more than a 90-day supply of any short-acting opioid medication(s) within the 12-month fiscal year 2010 (FY10) or any prescription for a long-acting opioid medication in FY10. The 90 days of supply criterion to define chronic, short-acting opioid use was established in consultation with an expert panel that included authors of the guideline [14]. Patients were excluded if they had a diagnosis of metastatic cancer. Short-acting opioids were defined as butorphanol, codeine, dihydrocodeine, hydrocodone, hydromorphone, meperidine, morphine, opium/belladonna, oxycodone, oxymorphone, pentazocine, and tapentadol. Long-acting opioids were defined as fentanyl, hydromorphone, methadone, sustained release morphine, sustained release oxycodone, sustained release oxymorphone, and sustained release tramadol. Data were obtained from VHA national administrative databases, including the VHA National Patient Care Database (NPCD), as well as VHA Decision Support System (DSS) pharmacy and laboratory files that were linked via scrambled social security numbers. This study was approved and overseen by the Stanford University Human Research Protection Program and the VA Palo Alto Health Care System Research and Development Committee. Metric Development The dependent variables in this project were developed to measure adherence across VHA facilities to recommendations in the 2010 Clinical Practice Guideline for the Management of Opioid Therapy for Chronic Pain (i.e., guideline-recommended practices; detailed descriptions and definitions of the variables are available in [14]). A panel of experts in opioid therapy and VHA pain management practice and policy met regularly for 1 year and guided the iterative development of individual metrics and metric definitions. These individuals are 113
Oliva et al. leading experts in the VHA on pain management, helped author the Clinical Practice Guideline, and had diverse professional and clinical backgrounds (e.g., pain medicine, pharmacy, behavioral medicine). A team of programmers with extensive experience with VHA administrative databases worked with the expert team throughout the development of the metrics to generate preliminary data and refine metrics as needed. These metrics were defined based on the following data elements: 1) dates of clinical encounters; 2) diagnoses treated (per International Classification of Diseases, 9th Revision, Clinical Modification [ICD-9] codes associated with encounters); 3) prescribed medications including fill date; 4) days’ supply; 5) formulation and daily dose; 6) labs ordered; 7) procedures conducted (per Current Procedural Terminology [CPT] and Healthcare Common Procedure Coding System [HCPCS] codes associated with encounters); and 8) location of care. Measures Demographic variables in this project were sex, age, marital status, and OEF/OIF status as coded in the FY10 VHA NPCD. Age was classified into three groups: 30 years old or younger, between 31 and 55 years old, and greater than 55 years old. Disorders were characterized based on ICD-9 diagnosis codes associated with VHA outpatient encounters or inpatient or residential stays as recorded in the NPCD. Any patient with at least one VHA health care visit for an ICD-9 diagnosis code in FY10 was characterized as positive for that disorder as follows: posttraumatic stress disorder (PTSD): 309.81; mood disorder (major depressive disorder or generalized anxiety disorder): 296.2–296.4 and 300.02; serious mental illness (SMI): 295, 297–299, 296.0–296.4, and 296.4 to 296.9 (affective psychosis, bipolar disorder, other psychosis, schizophrenia); and substance use disorder (SUD): 291–292.99, 303–305.1, 305.2–305.99 (alcohol use disorder, as well as the drug use disorders for opioids, cocaine, cannabis, amphetamines, barbiturates, and nicotine). Elixhauser comorbidity index categories of chronic pulmonary disease, cardiac arrhythmia, and liver disease (see Elixhauser et al. [15] for a listing of ICD-9 codes) were combined to create an index of medical frailty. These diagnoses were chosen specifically because they would be expected to influence opioid prescribing decisions. Metrics to Measure Adherence to GuidelineRecommended Practices A subset of the metrics described above was calculated for this study using FY10 data from the VHA NPCD and the VHA DSS dataset. The metrics measuring adherence to guideline-recommended practices included here are: 1. Ordering laboratory drug screens for (a) illicit drugs (e.g., cocaine, marijuana) and other illegal substances plus morphine-related opioid compounds (e.g., heroin), or (b) non-morphine opioids (e.g. hydrocodone, fentanyl). 2. Avoidance of sole reliance on opioids, as defined by receipt of (a) mental health assessment; (b) psycho114
therapy, including inpatient/outpatient and individual/ group treatment; (c) rehabilitation therapies including physical therapy, physical medicine and rehabilitation, recreational therapy, occupational therapy, pain clinic, or a prescribed exercise program; non-opioid pharmacotherapies including (d) anti-inflammatories, (e) tricyclic antidepressants (TCAs), (f) serotonin–norepinephrine reuptake inhibitors (SNRIs), and (g) anticonvulsants. 3. Side-effect management defined by prescription of a bowel regimen (i.e., a prescribed laxative or stool softener). 4. Pharmacy reconciliation defined by a pharmacy reconciliation encounter. 5. Contraindicated prescribing practices, as defined by receipt of (a) overlapping prescriptions for opioids and sedative medications (e.g., benzodiazepines, barbiturates, and carisoprodol [Soma]) and/or (b) single or overlapping prescriptions totaling more than 4 g per day of acetaminophen. Data Analysis Multilevel logistic regression was used to examine the two research questions with the primary independent variable being sex. Patients were linked to the VHA treatment facility where they received the majority of encounters or stays in FY10. The first set of analyses examined whether there were sex differences in receipt of the guidelinerecommended practices described above. Separate regressions were conducted with each of the guidelinerecommended practices included as dependent variables. Facility level rate of prescribing chronic, short-acting or long-acting opioids was included as a covariate and was defined as the percentage of total patients treated at the VHA facility who received a chronic, short-acting or longacting opioid prescription. The second set of analyses examined dependent variables that were identified as significant in the first set of analyses and, with sex as the independent variable, included the following covariates: 1) facility level rate of prescribing chronic, short-acting or long-acting opioids; 2) age category; 3) current marital status; 4) OEF/OIF status; 5) PTSD diagnosis; 6) SUD diagnosis; 7) mood disorder diagnosis; 8) SMI diagnosis; and 9) medical frailty. For any analyses with “yes” and “no” as the response, the reference group is “no.” Results Patient Characteristics In FY10, there were 480,809 VHA patients without metastatic cancer who received chronic, short-acting or long-acting prescription opioids. Of these patients, 389,335 (81%) received chronic, short-acting prescription opioids and 91,474 (19%) received long-acting prescription opioids. Table 1 describes the demographic characteristics of the sample. Overall, the sample was mostly male (93%), over age 55 (74%), married (53%), and diagnosed with a medical frailty (56%). Female VHA
Sex Differences in Pain Management Practices patients were significantly younger, more likely to be OEF/OIF veterans, less likely to be married, and had greater rates of PTSD, SMI, and mood disorder diagnoses, but lower rates of SUD diagnoses and medical frailty than male VHA patients (see Table 1).
Table 1
Primary Analyses The first set of multilevel logistic regression analyses revealed no significant sex differences between male and female VHA patients in prescription of a bowel regimen or in the use of urine drug screening for either illicit drugs or prescription opioids. For the remaining
Sample demographics by sex*
N Demographics Age >55 reference 31–55 30 OEF/OIF Currently married Yes Missing Diagnostic variables Medical frailty Posttraumatic stress disorder Substance use disorder Mood disorder Serious mental illness
Sample
Males
Females
480,809
449,006
31,803
P value
