bs_bs_banner
Pain Medicine 2014; 15: 1880–1891 Wiley Periodicals, Inc.
Prevalence of Pain Reporting and Associated Health Outcomes Across Emerging Markets and Developed Countries
Amir Goren, PhD,* Joaquín Mould-Quevedo, MSc, MBA, PhD,† and Marco daCosta DiBonaventura, PhD*
tics and measures of quality of life (SF-12v2 and SF-36v2), work productivity and activity impairment, and health care resource use.
*Health Outcomes Practice, Kantar Health; †EMEP Outcomes Research and Epidemiology, Pfizer, Inc., New York, New York, USA
Subjects. Respondents included 128,821 without pain and 29,848 with pain in developed countries, and 37,244 without pain and 4,789 with pain in emerging countries.
Reprint requests to: Amir Goren, PhD, Health Outcomes Practice, Kantar Health, 11 Madison Ave, 12th Floor, New York, NY 10010, USA. Tel: 1-212-706-3909; Fax: 1-212-647-7659; E-mail: [email protected]. Disclosure: Joaquín Mould-Quevedo was an employee of Pfizer, Inc. during the development of this manuscript. Amir Goren and Marco daCosta DiBonaventura are employees of Kantar Health, which conducted the National Health and Wellness Survey, and they analyzed the data on behalf of, and with funding from, Pfizer, Inc, including funding for the manuscript publication. All authors contributed to the study design, interpretation of results, and review, revision, and approval of the final manuscript.
Abstract Objective. The current study represents the first broad, multi-country, population-based survey of pain, assessing the association between pain and health outcomes, plus comparing the burden of pain across emerging and developed countries. Design. Data from the 2011/2012 National Health and Wellness Surveys were used. Respondents reporting pain (neuropathic pain, fibromyalgia, back pain, surgery pain, and/or arthritis pain) vs no pain in emerging (Brazil, China, Russia) vs developed (European Union, Japan, United States) countries were compared on sociodemographic characteris1880
Results. Pain reporting and treatment rates were lower in China (6.2% and 28.3%, respectively) and Japan (4.4% and 26.3%, respectively) than in other countries (≥14.3% and 35.8%, respectively). Significant impairments in quality of life, productivity, and resource use were associated with pain across all health outcomes in both developed and emerging countries, with some productivity and physical health status impairments greater with pain in developed countries, whereas mental health status impairment and resource use were greater with pain in emerging countries. Conclusions. Pain was associated with burden across all study outcomes in all regions. Yet, differences emerged in the degree of impairment, pain reporting, diagnosis, treatment rates, and characteristics of patients between emerging and developed nations, thus helping guide a broader understanding of this highly prevalent condition globally. Key Words. Pain Reporting; Prevalence; Treatment Prevalence; Health Outcomes; Emerging Markets; Developed Countries; Quality of Life
Introduction Pain, a highly prevalent condition globally [1], is generally defined as an “unpleasant sensory and emotional experience associated with actual or potential tissue damage” [2] and has been associated with burden across multiple
Pain Across Emerging and Developed Countries domains [3–7]. Chronic pain conditions constitute pain persisting beyond normal healing time [8] and can represent complex clinical challenges. Modalities of treatment may vary from classic pharmacologic interventions such as opioid analgesics, to more holistic approaches such as relaxation exercise. With opioids, addiction and treatment history have become important considerations and potential sources of undertreatment in severe pain management [9,10]. Recent reports on disease burden have found that musculoskeletal conditions (e.g., low back pain) are highly prevalent and a significant contributor to years lived with disability globally [11–14]. Low back pain specifically was found to be the leading cause of disability globally and in many regions, including the United States, Europe, and parts of Asia [1]. For example, between 1990 and 2010, a 57% increase in disability-adjusted life years was observed due to low back and neck pain in India [1]. Chronic pain conditions can be particularly burdensome and complex to treat and have been reported by up to 20% of survey samples [11,14–16]. Serious medical conditions such as cancer can be associated with even higher rates of pain, depending on the type of sample surveyed and disease staging [17]. Substantial costs are also associated with pain conditions, with a U.S.-based estimate of direct health care costs reported to be in excess of $261 billion, and indirect costs (e.g., lost productivity and absenteeism) estimated at an additional $300 billion [18]. Whereas a number of large-scale studies have examined the prevalence of pain conditions in the European Union (EU) and the United States [11,15,16,19–21], fewer studies exist outside of these regions. Research conducted in Russia has reported prevalence rates of over 30% for those experiencing low back pain [22], 13% for chronic generalized pain [23], and 46% for various types of arthralgia [24]. Research in Brazil has largely focused on pain reported by specific populations [25] or in conjunction with comorbid conditions [26]. A 2006 survey of adult employees found that nearly two-thirds of participants reported having a chronic pain condition, particularly among women [27], while a recent population-based survey found a high rate of back pain (12%) among respondents [28]. Similarly, few large-scale representative studies of pain have taken place throughout Asia, despite its emerging role on the world stage and its population density. In China, higher rates of pain have been reported in the northern regions of the country than in the south [29,30]. A review of the literature from 1983 to 1997 revealed 16 studies (two in English) and reported increased odds of suffering lower back pain when adopting certain physical positions (e.g., static posture) or being regularly exposed to low temperatures [31]. Studies of specific populations have reported rates of pain ranging from 38% to 50% [32–34]. In Japan, prevalence rates for musculoskeletal pain range between 15% and 41% [35,36].
Culture and national care models can exert an important influence on both the reporting and treatment of various pain conditions [37]. As noted, the majority of research examining pain and its correlates has been undertaken in the United States and Europe, while less research on the prevalence and treatment of pain has been published elsewhere in the world, particularly among emerging countries such as Brazil, China, or Russia. Little research therefore exists regarding possible differences between emerging and developed nations regarding prevalence rates of pain and its impact on broad measures of health and economic outcomes. The current study is intended to provide an important contribution to understanding pain prevalence and burden across developed and developing countries. The primary aim is to investigate the relationship between the experience of pain (neuropathic pain, fibromyalgia, back pain, pain from surgery or medical procedure, and/or arthritis pain) and health outcomes (healthrelated quality of life [HRQoL], work productivity loss, and health care resource utilization) in emerging (Brazil, China, Russia) and developed (EU, Japan, U.S.) nations. The secondary aim is to compare and contrast the burden of pain across emerging and developed countries and to test whether pain is associated with poorer outcomes in emerging vs developed countries. Methods Sample Data were provided from the 2011 and 2012 National Health and Wellness Surveys (NHWSs; Kantar Health, New York, NY, USA), an annual, cross-sectional, selfreport survey conducted with adults (≥18 years) in Brazil (2012), China (2012), Europe (EU: France, Germany, Italy, Spain, and UK) (2011), Japan (2012), Russia (2011), and the United States (2012). Stratified random sampling based on sex and age (at a minimum, depending on geography) was used to ensure NHWS representativeness of total regional adult populations. Invitations to participate were sent to members of the Lightspeed Research (LSR) Internet panel via e-mail and the survey was administered online, while some respondents over the age of 50 in Brazil, China, EU, and Russia were recruited in person or via phone and administered computer-assisted web interviews. The LSR panel comprises opt-in consumer panels, with members recruited via e-mail, e-newsletters, online banners, and coregistration with panel partners. Members receive points that can be accumulated and exchanged for prizes. The number of surveys in which members may participate per year is limited, and general response patterns are monitored, in order to ensure the quality of responses. Institutional review board (IRB) approval was granted by Essex IRB (Lebanon, NJ, USA), and the study complies with the principles of the Declaration of Helsinki [38]. NHWS U.S. demographic characteristics and weighted prevalence estimates of various comorbid conditions have compared favorably with corresponding results from the U.S. Census and the National Health Interview Survey [12,39,40]. Survey response rates varied as follows across 1881
Goren et al. geographies: Brazil (4.1%), China (16.1%), Europe (18.2%), Japan (18.4%), Russia (24.1%), and the United States (7.4%). Measures Pain Groups Respondents were divided according to whether or not they were experiencing pain. Patients who reported pain in the past 12 months and then attributed that pain (i.e., they were asked to indicate the conditions that caused them to feel pain in the past month) to either neuropathic pain, fibromyalgia, back pain, pain from surgery or medical procedure, and/or arthritis pain were classified as having pain. All others were classified as not having pain. Pain Characteristics Respondents self-reported the severity of their pain as mild, moderate, or severe. Respondents also indicated whether they currently used a prescription medication (or simply “medication” in Russia) to treat their pain.
Work Impairment Work impairment was assessed via the Work Productivity and Activity Impairment questionnaire, a validated measure of impairment due to health in the past 7 days that provides four subscales: absenteeism represents the percentage work time missed; presenteeism represents impairment while at work; overall work impairment represents the total work time impaired or missed; and activity impairment represents the percentage impairment of daily activities outside work [45]. Absenteeism, presenteeism, and overall work impairment were assessed only for respondents employed full-time, part-time, or selfemployed, whereas activity impairment was assessed for all respondents. Health care Resource Utilization Health care resource use was assessed via self-reported number of hospitalizations, emergency room (ER) visits, and provider visits in the past 6 months, as well as treatment (i.e., whether or not respondents were taking prescription medication for their pain condition). Analysis
Sociodemographics and Health History Age, sex, education (university or higher vs less), income (below country-specific median, above country-specific median, or decline to report), employment, body mass index (BMI) categories (normal weight, overweight, obese, or declined to provide height and weight vs underweight) were examined in bivariate and multivariable models described below. Also examined was the Charlson comorbidity index (CCI), which calculates respondents’ comorbid burden by weighting several comorbidities in terms of their contribution to mortality risk and then summing the result [41]. HRQoL HRQoL was measured with the Medical Outcomes Study Short Form (SF)-36v2 in China, Japan, and the United States, and the SF-12v2 in Brazil, EU, and Russia [42,43]. These questionnaires are multipurpose, generic health status instruments designed to assess eight domains of health (physical functioning, physical role limitations, bodily pain, general health, vitality, social functioning, emotional role limitations, and mental health). The SF-36v2 has 36 items, whereas the SF-12v2 is a brief, 12-item version. Better physical and mental health status are reflected in higher physical component summary and mental component summary (PCS and MCS) scores, respectively, while better overall HRQoL is reflected in higher SF-6D health utilities scores (theoretically ranging from 0 [death] to 1 [perfect health]), all derived from the SF-36v2 or SF-12v2 item responses [44]. PCS and MCS scores are normed to a mean of 50 and standard deviation of 10 for the U.S. population. HRQoL is assessed over the previous 4 weeks. 1882
Descriptive analyses were used to provide means/ standard deviations and percentages/frequencies for continuous and categorical variables, respectively. More specifically, descriptive analyses were used to show the prevalence of respondents experiencing pain and the proportion of those taking prescription medication for their pain. Those with pain were compared against all those without pain on their demographics, health characteristics, comorbidities, treatment, and health outcomes using t-tests and chi-square tests for continuous and categorical variables, respectively. Moreover, bivariates were used to compare pain and non-pain outcomes across emerging markets vs developed countries. Interactions were also examined, to assess whether the difference between pain and no pain was significantly larger in emerging or developed countries for each health outcome. This was done by entering the pain variable (pain vs no pain), the country type variable (emerging vs developed), and the interaction between the two (pain variable × country type) into a general linear model (for HRQoL variables) or generalized linear model (for all other outcomes). No further covariates were included at this stage. Significant interactions (P < 0.05) were reported. Multivariable regression models were used to further examine pain vs non-pain and emerging vs developed markets as predictors of health outcome impairments, controlling for sociodemographic and health history covariates (age, sex, education, annual household income, employment, BMI, and the CCI). For HRQoL variables, which exhibited a normal distribution, general linear models were used. Adjusted means from these models were reported for each group using a least squares algorithm. For all other variables (work productivity and health care resource use variables), generalized
Pain Across Emerging and Developed Countries linear models were used because of the pronounced skew of these variables. Specifically, a negative binomial distribution was specified along with a log-link function. Adjusted means from these models were reported on their original metrics (as opposed to the log of the original metric) for each group, using a maximum likelihood algorithm.
Japan (26.3%). Severe pain was reported by the smallest proportion of respondents in Russia, China, and Japan and by the greatest proportion in the United States. Characteristics of Patients with Pain Examining respondents with pain in each country, several differences were observed. NHWS respondents with pain in emerging countries (i.e., Brazil, China, and Russia) were younger by approximately 10 years than those with pain in developed countries (i.e., EU, Japan, the United States; see Table 2). Those with pain in emerging countries (particularly China, and to a lesser extent, Brazil) tended to be of higher socioeconomic status than those with pain in developed countries, based on greater rates of having above median incomes, a greater proportion currently employed, and a higher percentage with university education or above.
Interaction terms (pain variable × country type) were also added as an additional predictor to assess whether the difference between pain and no pain was significantly larger in emerging or developed countries for each outcome, controlling for the covariates listed above. All analyses were conducted in SAS version 9.3 (SAS Institute, Cary, NC, USA). The cutoff for statistical significance was P < 0.05. Results Prevalence of Pain, Treatment, and Levels of Severity
For both developed and emerging countries, those with pain were less likely to be male or of normal weight and more likely to be obese, and they had more comorbidities (see Table 3). However, contrary to developed countries, where respondents with pain were more socioeconomically disadvantaged than those without pain (i.e., fewer with university education, above median income, or employed), those with pain vs no pain in emerging countries did not differ appreciably in their levels of education, income, and employment—that is, NHWS respondents in emerging vs developed countries were on the whole of higher socioeconomic status.
Frequency of pain was generally comparable across Europe, Russia, and the United States (18.3% to 23.8%) and slightly lower in Brazil (14.3%; see Table 1). In both China and Japan, however, the prevalence of reported pain was much lower (6.2% and 4.4%, respectively). Across all countries except Russia (where only 39.2% of patients reported having been diagnosed with their pain condition), two-thirds of patients or more reported having been diagnosed with pain. Treatment rates were highest (40.1% to 50.4%) in Russia, Europe, and the United States and lowest in China (28.3%) and
Table 1 Prevalence of pain, diagnosed pain, prescription medication use, and pain severity across countries Emerging Countries
Experiencing pain (of the total country population) Diagnosed (of those with pain) Treated with prescription medication (of those with pain) Mild pain in the past week Moderate pain in the past week Severe pain in the past week
Developed Countries
Brazil N = 12,000
China N = 19,994
Russia N = 10,039
EU N = 57,512
Japan N = 30,000
United States N = 71,157
N
%
N
N
%
N
%
N
N
%
1,719
14.3
1,231
1,839
18.3
11,598
20.2
1,325
4.4
16,925
23.8
1,339
77.9
796
64.76
721
39.2
9,350
80.6
848
64.0
13,108
77.4
615
35.8
348
28.3
927
50.4
5,445
47.0
349
26.3
6,790
40.1
605
35.2
542
44.0
920
50.0
3,556
30.7
538
40.6
4,646
27.5
558
32.5
443
36.0
599
32.6
4,140
35.7
471
35.5
5,878
34.7
556
32.3
246
20.0
320
17.4
3,902
33.6
316
23.8
6,401
37.8
% 6.2
%
EU = European Union.
1883
1884 473 (38.4) 742 (60.3) 16 (1.3) 1,012 (82.2) 97 (7.9) 786 (63.9) 238 (19.3) 102 (8.3) 8 (0.6) 0.80 ± 1.94
750 (43.6) 787 (45.8) 182 (10.6) 1,130 (65.7) 64 (3.7) 628 (36.5) 537 (31.2) 479 (27.9) 11 (0.6) 0.55 ± 1.30
P values represent differences across all six countries/regions. EU = European Union; SD = standard deviation.
41.26 ± 13.46 618 (50.2) 752 (61.1)
38.99 ± 13.24 690 (40.1) 625 (36.4)
0.77 ± 1.60
94 (5.1) 833 (45.3) 540 (29.4) 366 (19.9) 6 (0.3)
759 (41.3) 1,034 (56.2) 46 (2.5) 1,340 (72.9)
41.07 ± 14.44 791 (43.0) 1,148 (62.4)
0.49 ± 1.07
273 (2.4) 4,186 (36.1) 3,843 (33.1) 3,016 (26.0) 280 (2.4)
5,881 (50.7) 4,241 (36.6) 1,476 (12.7) 5,834 (50.3)
48.34 ± 14.94 4,871 (42.0) 4,253 (36.7)
0.39 ± 1.06
162 (12.2) 858 (64.8) 202 (15.2) 59 (4.5) 44 (3.3)
657 (49.6) 570 (43.0) 98 (7.4) 760 (57.4)
49.46 ± 14.22 580 (43.8) 532 (40.2)
Japan (N = 1,325)
EU (N = 11,598)
Russia (N = 1,839)
Brazil (N = 1,719)
China (N = 1,231)
Developed Countries
Emerging Countries
Sociodemographic and health history variables of patients with pain in each country
Age (years) Mean ± SD Male (%) University education or higher (%) Annual household income Below country median (%) Above country median (%) Decline to answer (%) Employed (%) Body mass index (BMI) category Underweight (%) Normal weight (%) Overweight (%) Obese (%) Decline to provide weight (%) Charlson comorbidity index Mean ± SD
Table 2
0.79 ± 1.29
253 (1.5) 3,935 (23.2) 5,064 (29.9) 7,308 (43.2) 365 (2.2)
9,051 (53.5) 6,684 (39.5) 1,190 (7.0) 7,066 (41.7)
52.54 ± 15.89 7,921 (46.8) 5,479 (32.4)
United States (N = 16,925)
Pain Medicine 2014; 15: 1880–1891 Wiley Periodicals, Inc.
Prevalence of Pain Reporting and Associated Health Outcomes Across Emerging Markets and Developed Countries
Amir Goren, PhD,* Joaquín Mould-Quevedo, MSc, MBA, PhD,† and Marco daCosta DiBonaventura, PhD*
tics and measures of quality of life (SF-12v2 and SF-36v2), work productivity and activity impairment, and health care resource use.
*Health Outcomes Practice, Kantar Health; †EMEP Outcomes Research and Epidemiology, Pfizer, Inc., New York, New York, USA
Subjects. Respondents included 128,821 without pain and 29,848 with pain in developed countries, and 37,244 without pain and 4,789 with pain in emerging countries.
Reprint requests to: Amir Goren, PhD, Health Outcomes Practice, Kantar Health, 11 Madison Ave, 12th Floor, New York, NY 10010, USA. Tel: 1-212-706-3909; Fax: 1-212-647-7659; E-mail: [email protected]. Disclosure: Joaquín Mould-Quevedo was an employee of Pfizer, Inc. during the development of this manuscript. Amir Goren and Marco daCosta DiBonaventura are employees of Kantar Health, which conducted the National Health and Wellness Survey, and they analyzed the data on behalf of, and with funding from, Pfizer, Inc, including funding for the manuscript publication. All authors contributed to the study design, interpretation of results, and review, revision, and approval of the final manuscript.
Abstract Objective. The current study represents the first broad, multi-country, population-based survey of pain, assessing the association between pain and health outcomes, plus comparing the burden of pain across emerging and developed countries. Design. Data from the 2011/2012 National Health and Wellness Surveys were used. Respondents reporting pain (neuropathic pain, fibromyalgia, back pain, surgery pain, and/or arthritis pain) vs no pain in emerging (Brazil, China, Russia) vs developed (European Union, Japan, United States) countries were compared on sociodemographic characteris1880
Results. Pain reporting and treatment rates were lower in China (6.2% and 28.3%, respectively) and Japan (4.4% and 26.3%, respectively) than in other countries (≥14.3% and 35.8%, respectively). Significant impairments in quality of life, productivity, and resource use were associated with pain across all health outcomes in both developed and emerging countries, with some productivity and physical health status impairments greater with pain in developed countries, whereas mental health status impairment and resource use were greater with pain in emerging countries. Conclusions. Pain was associated with burden across all study outcomes in all regions. Yet, differences emerged in the degree of impairment, pain reporting, diagnosis, treatment rates, and characteristics of patients between emerging and developed nations, thus helping guide a broader understanding of this highly prevalent condition globally. Key Words. Pain Reporting; Prevalence; Treatment Prevalence; Health Outcomes; Emerging Markets; Developed Countries; Quality of Life
Introduction Pain, a highly prevalent condition globally [1], is generally defined as an “unpleasant sensory and emotional experience associated with actual or potential tissue damage” [2] and has been associated with burden across multiple
Pain Across Emerging and Developed Countries domains [3–7]. Chronic pain conditions constitute pain persisting beyond normal healing time [8] and can represent complex clinical challenges. Modalities of treatment may vary from classic pharmacologic interventions such as opioid analgesics, to more holistic approaches such as relaxation exercise. With opioids, addiction and treatment history have become important considerations and potential sources of undertreatment in severe pain management [9,10]. Recent reports on disease burden have found that musculoskeletal conditions (e.g., low back pain) are highly prevalent and a significant contributor to years lived with disability globally [11–14]. Low back pain specifically was found to be the leading cause of disability globally and in many regions, including the United States, Europe, and parts of Asia [1]. For example, between 1990 and 2010, a 57% increase in disability-adjusted life years was observed due to low back and neck pain in India [1]. Chronic pain conditions can be particularly burdensome and complex to treat and have been reported by up to 20% of survey samples [11,14–16]. Serious medical conditions such as cancer can be associated with even higher rates of pain, depending on the type of sample surveyed and disease staging [17]. Substantial costs are also associated with pain conditions, with a U.S.-based estimate of direct health care costs reported to be in excess of $261 billion, and indirect costs (e.g., lost productivity and absenteeism) estimated at an additional $300 billion [18]. Whereas a number of large-scale studies have examined the prevalence of pain conditions in the European Union (EU) and the United States [11,15,16,19–21], fewer studies exist outside of these regions. Research conducted in Russia has reported prevalence rates of over 30% for those experiencing low back pain [22], 13% for chronic generalized pain [23], and 46% for various types of arthralgia [24]. Research in Brazil has largely focused on pain reported by specific populations [25] or in conjunction with comorbid conditions [26]. A 2006 survey of adult employees found that nearly two-thirds of participants reported having a chronic pain condition, particularly among women [27], while a recent population-based survey found a high rate of back pain (12%) among respondents [28]. Similarly, few large-scale representative studies of pain have taken place throughout Asia, despite its emerging role on the world stage and its population density. In China, higher rates of pain have been reported in the northern regions of the country than in the south [29,30]. A review of the literature from 1983 to 1997 revealed 16 studies (two in English) and reported increased odds of suffering lower back pain when adopting certain physical positions (e.g., static posture) or being regularly exposed to low temperatures [31]. Studies of specific populations have reported rates of pain ranging from 38% to 50% [32–34]. In Japan, prevalence rates for musculoskeletal pain range between 15% and 41% [35,36].
Culture and national care models can exert an important influence on both the reporting and treatment of various pain conditions [37]. As noted, the majority of research examining pain and its correlates has been undertaken in the United States and Europe, while less research on the prevalence and treatment of pain has been published elsewhere in the world, particularly among emerging countries such as Brazil, China, or Russia. Little research therefore exists regarding possible differences between emerging and developed nations regarding prevalence rates of pain and its impact on broad measures of health and economic outcomes. The current study is intended to provide an important contribution to understanding pain prevalence and burden across developed and developing countries. The primary aim is to investigate the relationship between the experience of pain (neuropathic pain, fibromyalgia, back pain, pain from surgery or medical procedure, and/or arthritis pain) and health outcomes (healthrelated quality of life [HRQoL], work productivity loss, and health care resource utilization) in emerging (Brazil, China, Russia) and developed (EU, Japan, U.S.) nations. The secondary aim is to compare and contrast the burden of pain across emerging and developed countries and to test whether pain is associated with poorer outcomes in emerging vs developed countries. Methods Sample Data were provided from the 2011 and 2012 National Health and Wellness Surveys (NHWSs; Kantar Health, New York, NY, USA), an annual, cross-sectional, selfreport survey conducted with adults (≥18 years) in Brazil (2012), China (2012), Europe (EU: France, Germany, Italy, Spain, and UK) (2011), Japan (2012), Russia (2011), and the United States (2012). Stratified random sampling based on sex and age (at a minimum, depending on geography) was used to ensure NHWS representativeness of total regional adult populations. Invitations to participate were sent to members of the Lightspeed Research (LSR) Internet panel via e-mail and the survey was administered online, while some respondents over the age of 50 in Brazil, China, EU, and Russia were recruited in person or via phone and administered computer-assisted web interviews. The LSR panel comprises opt-in consumer panels, with members recruited via e-mail, e-newsletters, online banners, and coregistration with panel partners. Members receive points that can be accumulated and exchanged for prizes. The number of surveys in which members may participate per year is limited, and general response patterns are monitored, in order to ensure the quality of responses. Institutional review board (IRB) approval was granted by Essex IRB (Lebanon, NJ, USA), and the study complies with the principles of the Declaration of Helsinki [38]. NHWS U.S. demographic characteristics and weighted prevalence estimates of various comorbid conditions have compared favorably with corresponding results from the U.S. Census and the National Health Interview Survey [12,39,40]. Survey response rates varied as follows across 1881
Goren et al. geographies: Brazil (4.1%), China (16.1%), Europe (18.2%), Japan (18.4%), Russia (24.1%), and the United States (7.4%). Measures Pain Groups Respondents were divided according to whether or not they were experiencing pain. Patients who reported pain in the past 12 months and then attributed that pain (i.e., they were asked to indicate the conditions that caused them to feel pain in the past month) to either neuropathic pain, fibromyalgia, back pain, pain from surgery or medical procedure, and/or arthritis pain were classified as having pain. All others were classified as not having pain. Pain Characteristics Respondents self-reported the severity of their pain as mild, moderate, or severe. Respondents also indicated whether they currently used a prescription medication (or simply “medication” in Russia) to treat their pain.
Work Impairment Work impairment was assessed via the Work Productivity and Activity Impairment questionnaire, a validated measure of impairment due to health in the past 7 days that provides four subscales: absenteeism represents the percentage work time missed; presenteeism represents impairment while at work; overall work impairment represents the total work time impaired or missed; and activity impairment represents the percentage impairment of daily activities outside work [45]. Absenteeism, presenteeism, and overall work impairment were assessed only for respondents employed full-time, part-time, or selfemployed, whereas activity impairment was assessed for all respondents. Health care Resource Utilization Health care resource use was assessed via self-reported number of hospitalizations, emergency room (ER) visits, and provider visits in the past 6 months, as well as treatment (i.e., whether or not respondents were taking prescription medication for their pain condition). Analysis
Sociodemographics and Health History Age, sex, education (university or higher vs less), income (below country-specific median, above country-specific median, or decline to report), employment, body mass index (BMI) categories (normal weight, overweight, obese, or declined to provide height and weight vs underweight) were examined in bivariate and multivariable models described below. Also examined was the Charlson comorbidity index (CCI), which calculates respondents’ comorbid burden by weighting several comorbidities in terms of their contribution to mortality risk and then summing the result [41]. HRQoL HRQoL was measured with the Medical Outcomes Study Short Form (SF)-36v2 in China, Japan, and the United States, and the SF-12v2 in Brazil, EU, and Russia [42,43]. These questionnaires are multipurpose, generic health status instruments designed to assess eight domains of health (physical functioning, physical role limitations, bodily pain, general health, vitality, social functioning, emotional role limitations, and mental health). The SF-36v2 has 36 items, whereas the SF-12v2 is a brief, 12-item version. Better physical and mental health status are reflected in higher physical component summary and mental component summary (PCS and MCS) scores, respectively, while better overall HRQoL is reflected in higher SF-6D health utilities scores (theoretically ranging from 0 [death] to 1 [perfect health]), all derived from the SF-36v2 or SF-12v2 item responses [44]. PCS and MCS scores are normed to a mean of 50 and standard deviation of 10 for the U.S. population. HRQoL is assessed over the previous 4 weeks. 1882
Descriptive analyses were used to provide means/ standard deviations and percentages/frequencies for continuous and categorical variables, respectively. More specifically, descriptive analyses were used to show the prevalence of respondents experiencing pain and the proportion of those taking prescription medication for their pain. Those with pain were compared against all those without pain on their demographics, health characteristics, comorbidities, treatment, and health outcomes using t-tests and chi-square tests for continuous and categorical variables, respectively. Moreover, bivariates were used to compare pain and non-pain outcomes across emerging markets vs developed countries. Interactions were also examined, to assess whether the difference between pain and no pain was significantly larger in emerging or developed countries for each health outcome. This was done by entering the pain variable (pain vs no pain), the country type variable (emerging vs developed), and the interaction between the two (pain variable × country type) into a general linear model (for HRQoL variables) or generalized linear model (for all other outcomes). No further covariates were included at this stage. Significant interactions (P < 0.05) were reported. Multivariable regression models were used to further examine pain vs non-pain and emerging vs developed markets as predictors of health outcome impairments, controlling for sociodemographic and health history covariates (age, sex, education, annual household income, employment, BMI, and the CCI). For HRQoL variables, which exhibited a normal distribution, general linear models were used. Adjusted means from these models were reported for each group using a least squares algorithm. For all other variables (work productivity and health care resource use variables), generalized
Pain Across Emerging and Developed Countries linear models were used because of the pronounced skew of these variables. Specifically, a negative binomial distribution was specified along with a log-link function. Adjusted means from these models were reported on their original metrics (as opposed to the log of the original metric) for each group, using a maximum likelihood algorithm.
Japan (26.3%). Severe pain was reported by the smallest proportion of respondents in Russia, China, and Japan and by the greatest proportion in the United States. Characteristics of Patients with Pain Examining respondents with pain in each country, several differences were observed. NHWS respondents with pain in emerging countries (i.e., Brazil, China, and Russia) were younger by approximately 10 years than those with pain in developed countries (i.e., EU, Japan, the United States; see Table 2). Those with pain in emerging countries (particularly China, and to a lesser extent, Brazil) tended to be of higher socioeconomic status than those with pain in developed countries, based on greater rates of having above median incomes, a greater proportion currently employed, and a higher percentage with university education or above.
Interaction terms (pain variable × country type) were also added as an additional predictor to assess whether the difference between pain and no pain was significantly larger in emerging or developed countries for each outcome, controlling for the covariates listed above. All analyses were conducted in SAS version 9.3 (SAS Institute, Cary, NC, USA). The cutoff for statistical significance was P < 0.05. Results Prevalence of Pain, Treatment, and Levels of Severity
For both developed and emerging countries, those with pain were less likely to be male or of normal weight and more likely to be obese, and they had more comorbidities (see Table 3). However, contrary to developed countries, where respondents with pain were more socioeconomically disadvantaged than those without pain (i.e., fewer with university education, above median income, or employed), those with pain vs no pain in emerging countries did not differ appreciably in their levels of education, income, and employment—that is, NHWS respondents in emerging vs developed countries were on the whole of higher socioeconomic status.
Frequency of pain was generally comparable across Europe, Russia, and the United States (18.3% to 23.8%) and slightly lower in Brazil (14.3%; see Table 1). In both China and Japan, however, the prevalence of reported pain was much lower (6.2% and 4.4%, respectively). Across all countries except Russia (where only 39.2% of patients reported having been diagnosed with their pain condition), two-thirds of patients or more reported having been diagnosed with pain. Treatment rates were highest (40.1% to 50.4%) in Russia, Europe, and the United States and lowest in China (28.3%) and
Table 1 Prevalence of pain, diagnosed pain, prescription medication use, and pain severity across countries Emerging Countries
Experiencing pain (of the total country population) Diagnosed (of those with pain) Treated with prescription medication (of those with pain) Mild pain in the past week Moderate pain in the past week Severe pain in the past week
Developed Countries
Brazil N = 12,000
China N = 19,994
Russia N = 10,039
EU N = 57,512
Japan N = 30,000
United States N = 71,157
N
%
N
N
%
N
%
N
N
%
1,719
14.3
1,231
1,839
18.3
11,598
20.2
1,325
4.4
16,925
23.8
1,339
77.9
796
64.76
721
39.2
9,350
80.6
848
64.0
13,108
77.4
615
35.8
348
28.3
927
50.4
5,445
47.0
349
26.3
6,790
40.1
605
35.2
542
44.0
920
50.0
3,556
30.7
538
40.6
4,646
27.5
558
32.5
443
36.0
599
32.6
4,140
35.7
471
35.5
5,878
34.7
556
32.3
246
20.0
320
17.4
3,902
33.6
316
23.8
6,401
37.8
% 6.2
%
EU = European Union.
1883
1884 473 (38.4) 742 (60.3) 16 (1.3) 1,012 (82.2) 97 (7.9) 786 (63.9) 238 (19.3) 102 (8.3) 8 (0.6) 0.80 ± 1.94
750 (43.6) 787 (45.8) 182 (10.6) 1,130 (65.7) 64 (3.7) 628 (36.5) 537 (31.2) 479 (27.9) 11 (0.6) 0.55 ± 1.30
P values represent differences across all six countries/regions. EU = European Union; SD = standard deviation.
41.26 ± 13.46 618 (50.2) 752 (61.1)
38.99 ± 13.24 690 (40.1) 625 (36.4)
0.77 ± 1.60
94 (5.1) 833 (45.3) 540 (29.4) 366 (19.9) 6 (0.3)
759 (41.3) 1,034 (56.2) 46 (2.5) 1,340 (72.9)
41.07 ± 14.44 791 (43.0) 1,148 (62.4)
0.49 ± 1.07
273 (2.4) 4,186 (36.1) 3,843 (33.1) 3,016 (26.0) 280 (2.4)
5,881 (50.7) 4,241 (36.6) 1,476 (12.7) 5,834 (50.3)
48.34 ± 14.94 4,871 (42.0) 4,253 (36.7)
0.39 ± 1.06
162 (12.2) 858 (64.8) 202 (15.2) 59 (4.5) 44 (3.3)
657 (49.6) 570 (43.0) 98 (7.4) 760 (57.4)
49.46 ± 14.22 580 (43.8) 532 (40.2)
Japan (N = 1,325)
EU (N = 11,598)
Russia (N = 1,839)
Brazil (N = 1,719)
China (N = 1,231)
Developed Countries
Emerging Countries
Sociodemographic and health history variables of patients with pain in each country
Age (years) Mean ± SD Male (%) University education or higher (%) Annual household income Below country median (%) Above country median (%) Decline to answer (%) Employed (%) Body mass index (BMI) category Underweight (%) Normal weight (%) Overweight (%) Obese (%) Decline to provide weight (%) Charlson comorbidity index Mean ± SD
Table 2
0.79 ± 1.29
253 (1.5) 3,935 (23.2) 5,064 (29.9) 7,308 (43.2) 365 (2.2)
9,051 (53.5) 6,684 (39.5) 1,190 (7.0) 7,066 (41.7)
52.54 ± 15.89 7,921 (46.8) 5,479 (32.4)
United States (N = 16,925)
