The Journal of Pain, Vol 15, No 11 (November), 2014: pp 1081-1091 Available online at www.jpain.org and www.sciencedirect.com
Original Reports Chronic Pain Prevalence and Associated Factors in a Segment ~o Paulo City of the Population of Sa Dayane Maia Costa Cabral, Eduardo Sawaya Botelho Bracher, Jidiene Dylese Prescatan Depintor, and Jos e Eluf-Neto ~o Paulo School of Medicine, Sa ~o Paulo, Brazil. University of Sa
Abstract: A cross-sectional epidemiologic survey was performed to determine the prevalence of chronic pain (CP) and to identify associated factors in a random sample of persons 15 years or older ~o Paulo City, Brazil. A total of 1,108 eligible participants were from a segment of the population of Sa randomly selected, and face-to-face interviews were performed with 826 individuals (74.5%) between December 2011 and February 2012. Chronic Pain Grade, Hospital Anxiety and Depression Scale, and EuroQol-5D were used to verify pain characteristics and the associated signs of psychological distress. A prevalence of 42% (95% confidence interval, 38.6–45.4) was observed for CP, and the participants with CP had an average pain intensity of 5.9 (standard deviation = 1.9) and a pain-related disability of 4.1 (standard deviation = 3.2) on a 0 to 10 scale. Persistent pain was present in 68.6% of those with CP, and 32.8% of the population sample had high-intensity or high-interference pain (Chronic Pain Grade II, III, and IV). Quality of life was significantly worse among the CP individuals. The following factors were independently associated with CP: female gender, age 30 years or older, #4 years of education, symptoms consistent with anxiety, and intense physical strain. Indicators of pain severity increased with pain grades. Perspective: CP is highly prevalent in the city of Sa~o Paulo and has a considerable impact on healthrelated quality of life. Demographic, socioeconomic, and psychological factors are independently associated with this condition. ª 2014 by the American Pain Society Key words: Chronic pain, prevalence, cross-sectional studies, disability, quality of life.
C
hronic pain (CP) is defined as pain that persists past the normal time of healing; for nonmalignant pain, 3 months has been suggested as the most convenient point of division between acute and chronic pain.45 CP is commonly regarded as a multidimensional phenomenon that involves physical, psychological, and sociocultural aspects3,11,29 and impacts the individual’s health and well-being, health care services, and society
Received February 5, 2014; Revised June 24, 2014; Accepted July 1, 2014. ~o Paulo, FacSupported by Faculdade de Medicina da Universidade de Sa rdia de Sa ~o Paulo and Cenuldade de Medicina da Santa Casa de Miserico ~o Paulo. de Escola Barra Funda, Sa tro de Sau The authors declare no conflicts of interest. Address reprint requests to Dayane Maia Costa Cabral, DC, MSc, ~o Paulo School of Medicine, Sa ~o Paulo, Brazil. E-mail: University of Sa [email protected] 1526-5900/$36.00 ª 2014 by the American Pain Society http://dx.doi.org/10.1016/j.jpain.2014.07.001
as a whole.64 Musculoskeletal disorders have recently been listed as the first cause of ‘‘years lived with disability’’ worldwide49 and are one of the most common reasons for which people seek medical care.12 Estimates of the prevalence of CP range from 10.1 to 55.2%, and a weighted mean prevalence of 30.3% has been suggested.24,31 Cultural aspects and methodologic differences are commonly mentioned as explanations for the wide variation observed across studies.31,33,67 It has been suggested that CP should be considered a distinct diagnostic entity and that genetic, psychological, and social factors contribute to the perception and expression of the condition.59,63 Health measurement instruments are an efficient tool for the standardized identification of attributes of pain, including intensity, disability, persistence, and associated behavioral factors.29,62,69,70 Additionally, the use of previously validated scales may provide a common ground for comparison between studies. 1081
1082
The Journal of Pain
Several factors have been associated with CP. The condition is most commonly reported by females and older persons.6,8,25,33,55,57,67,68 The associated social conditions include unemployment, a lower socioeconomic level, a lower educational level, and employment that requires high physical strain.4,6,8,25,33 Depression and smoking are 2 psychosocial characteristics mentioned.2,4,11,15,21,51,57 Other associated factors include obesity, marital status (divorced or separated), and poor self-rated health.8,25 The current epidemiologic survey on the prevalence of CP in the most populated city in Latin America may provide useful data for the development of effective health care policies in a region where few populationbased studies on this subject have been published.57,68 The discrimination of specific groups according to the severity of symptoms might aid in targeting interventions to persons at higher risk. The present study was conducted to estimate the prevalence of CP in a ~o segment of the general population of the city of Sa Paulo and to identify sociodemographic, psychosocial, and occupational factors associated with CP.
Methods Sampling and Procedures A cross-sectional, population-based epidemiologic ~o Paulo City study was conducted in a central part of Sa under the administration of the Brazilian Family Health Program. All of the families that resided in this area were registered at a specific Health Care Unit in the district of Barra Funda. The community comprised 8,052 individuals from 2,549 families. The households were divided into 17 microareas, each comprising a geographically delimited area with approximately 150 families. A community health agent was responsible for visiting the families from each microarea once per month. Population changes noted during those visits were used to update the records, thus providing a reasonably reliable registry. The current study was approved by the ethics and research committees of Faculdade de Medicina da Universidade de ~o Paulo, Faculdade de Medicina da Santa Casa de MiserSa rdia de Sa ~o Paulo, and the Sa ~o Paulo City Prefecture. ico Participants provided informed consent. Permission was also obtained for the use of the EuroQol-5D. A CP prevalence of 25% was expected based on a ~o recent study that was conducted in the city of Sa Paulo.10 A sample size of 820 participants was estimated, considering a sampling error of 3% and a confidence interval of 95%. The participants were selected via random probability sampling. The number of families selected from each microarea was determined in proportion to the relative size of the microarea. Then, the families, identified by numbers, were randomly selected (Excel Sample Test, 2010.11; Microsoft Inc, Redmond, WA). One individual from each family was chosen from a preordained table of family members listed according to age and gender in relation to the head of the household.34 The inclusion criteria were persons of either gender aged 15 or older. Persons who were unable to participate in the interview
Chronic Pain Prevalence in S~ ao Paulo City for any reason and those who were considered to be dangerous to the interviewers were excluded. Each household was visited at least 4 times at different hours and on different days of the week in an attempt to maximize the chance of locating the person. If contact was not established by the fourth attempt, a new individual from a different family was selected using the same method.
Instruments and Data Collection The data were collected through domiciliary interviews that were conducted by previously trained university students. A pilot study with 30 subjects was conducted at a university ambulatory clinic to train the interviewers and identify the subjects’ perceived difficulties in response to the questionnaire. The interviews consisted of an initial questionnaire, followed by the application of 4 health measurement scales and 1 socioeconomic classification instrument. To gather reliable standardized information on factors associated with CP, validated instruments were used. CP was measured using the Chronic Pain Grade (CPG), an 8-item questionnaire that evaluates the most bothersome pain in the preceding 3 months.69 This instrument classifies CP into 4 hierarchical and mutually exclusive grades. The pain intensity is calculated as the average of 3 questions concerning the present, worst, and average pain, as measured on a 0 to 10 scale, ranging from ‘‘no pain’’ to ‘‘pain as bad as could be.’’ ‘‘Disability days’’ refers to the number of days for which pain prevented usual activities. The pain-related disability is calculated as the average of 3 questions on pain interference with daily activities, social and family activities, and work, as measured on a 0 to 10 scale, ranging from ‘‘no interference’’ to ‘‘unable to carry on any activities.’’ The pain grade is assigned to the following 4 hierarchical and mutually exclusive classes: I, low intensity; II, high intensity; III, moderately limiting; and IV, severely limiting. The psychometric properties of the CPG have been studied in different population settings,9,23,69 and the CGP has been used in epidemiologic surveys,23,41,54,74 clinical trials,66 and studies on risk factors for chronic pain.15,44 Health measurement scales with variables that are known to be associated with CP were also included in the present study. All instruments were previously culturally adapted to Brazilian Portuguese, and validation studies were performed on the Brazilian population. The Hospital Anxiety and Depression Scale (HADS) was developed to identify symptoms that are associated with anxiety and depression in nonpsychiatric individuals.7,16,75 This scale is divided into anxiety and depression subscales, each consisting of 7 intermingled multiple-choice questions. Four alternatives, weighed 0 to 3, are provided for each question. A score above 8 is considered the optimal threshold for anxiety and depression, with a sensitivity and specificity of approximately .80 for both subscales. The HADS has been extensively used and is frequently applied in population studies. Furthermore, its psychometric properties have been studied in several populations.7,54 In the current study, anxiety and depression were considered to be
Maia Costa Cabral et al present with a score equal to or greater than 9 points on the respective subscale. The Alcohol Use Disorders Identification Test was developed by the World Health Organization as a screening instrument for excessive alcohol use. It consists of 10 questions that evaluate the recent use of alcohol, symptoms of dependence, and problems related to alcohol use. Respondents are classified into 1 of the following 3 domains: hazardous alcohol use, dependence symptoms, and harmful alcohol use.39,42 € m Test for Nicotine Dependence is a The Fagerstro screening tool that provides a self-reported measure of nicotine dependence. It consists of an 8-item scale that provides 5 grades of dependence, ranging from very low to very high.14,43 The EuroQol-5D was developed by the EuroQol group as an instrument for the self-assessment of healthrelated quality of life.19,26 It consists of 5 multiplechoice questions that provide scores on the following 5 dimensions: mobility, self-care, usual activities, pain/ discomfort, and anxiety/depression. The final score, known as the EuroIndex, ranges from 0 (the lowest score and worst health status) to 1 (the highest score and best health status). Additionally, the global rating of current health, as assessed by a numeric scale that ranges from 0 (the worse health state) to 100 (the best health state), is commonly used as an indicator of self-rated health. The psychometric properties of the EuroQol-5D have been studied in several populations.38,40 The Brazil Economic Classification Criteria classifies the population into 5 economic segments, ranging from A to E, according to family income.20 During the interviews, the demographic characteristics, pain symptoms, number of days of pain medication use during the past 30 days, and number of pain-related medical visits during the previous 3 months were recorded. CP was considered present when participants reported pain of any intensity for 3 months or more and at least 1 episode of pain during the past 30 days. Persons with CP were asked to draw their pain on a body map of the human figure and to signal with an arrow the site of the most bothersome pain. CPG was then completed based on the location of that pain. Chronic widespread pain was defined according to the criteria of the American College of Rheumatology as pain that is present below and above the waistline and on both sides of the body.73 All participants completed the HADS, Alcohol Use Disorders Identification Test, and Brazil Economic Classification Criteria questionnaires; those who smoked also re€ m Test for Nicotine Dependence. sponded to the Fagerstro
Statistical Analysis Statistical analysis was performed using Stata, 12.0 (StataCorp LP, College Station, TX). Descriptive analyses of the median, mean, standard deviation, and percentage were used to establish the demographic and clinical characteristics of the sample. The CP prevalence and its 95% confidence interval (CI) were determined. The association between CP and the selected variables was estimated by prevalence ratios and 95% CI. Initial
The Journal of Pain
1083
analysis included the chi-square test. For ordered (categorical) variables, the chi-square test for a linear trend in the prevalence ratio was used by categorizing the exposure variables and entering the continuous scores. Variables with a P value less than .20 in univariate analysis were selected for a multiple regression analysis. We used a Cox regression model with constant time and robust variance.5,36,37 Usually, Cox regression is employed to analyze time-to-event data. When a constant risk period (time = 1) is assigned to all subjects, the hazard rate ratio estimated by Cox regression equals the prevalence ratio in cross-sectional studies.5,36 The final model included only those variables with P
Original Reports Chronic Pain Prevalence and Associated Factors in a Segment ~o Paulo City of the Population of Sa Dayane Maia Costa Cabral, Eduardo Sawaya Botelho Bracher, Jidiene Dylese Prescatan Depintor, and Jos e Eluf-Neto ~o Paulo School of Medicine, Sa ~o Paulo, Brazil. University of Sa
Abstract: A cross-sectional epidemiologic survey was performed to determine the prevalence of chronic pain (CP) and to identify associated factors in a random sample of persons 15 years or older ~o Paulo City, Brazil. A total of 1,108 eligible participants were from a segment of the population of Sa randomly selected, and face-to-face interviews were performed with 826 individuals (74.5%) between December 2011 and February 2012. Chronic Pain Grade, Hospital Anxiety and Depression Scale, and EuroQol-5D were used to verify pain characteristics and the associated signs of psychological distress. A prevalence of 42% (95% confidence interval, 38.6–45.4) was observed for CP, and the participants with CP had an average pain intensity of 5.9 (standard deviation = 1.9) and a pain-related disability of 4.1 (standard deviation = 3.2) on a 0 to 10 scale. Persistent pain was present in 68.6% of those with CP, and 32.8% of the population sample had high-intensity or high-interference pain (Chronic Pain Grade II, III, and IV). Quality of life was significantly worse among the CP individuals. The following factors were independently associated with CP: female gender, age 30 years or older, #4 years of education, symptoms consistent with anxiety, and intense physical strain. Indicators of pain severity increased with pain grades. Perspective: CP is highly prevalent in the city of Sa~o Paulo and has a considerable impact on healthrelated quality of life. Demographic, socioeconomic, and psychological factors are independently associated with this condition. ª 2014 by the American Pain Society Key words: Chronic pain, prevalence, cross-sectional studies, disability, quality of life.
C
hronic pain (CP) is defined as pain that persists past the normal time of healing; for nonmalignant pain, 3 months has been suggested as the most convenient point of division between acute and chronic pain.45 CP is commonly regarded as a multidimensional phenomenon that involves physical, psychological, and sociocultural aspects3,11,29 and impacts the individual’s health and well-being, health care services, and society
Received February 5, 2014; Revised June 24, 2014; Accepted July 1, 2014. ~o Paulo, FacSupported by Faculdade de Medicina da Universidade de Sa rdia de Sa ~o Paulo and Cenuldade de Medicina da Santa Casa de Miserico ~o Paulo. de Escola Barra Funda, Sa tro de Sau The authors declare no conflicts of interest. Address reprint requests to Dayane Maia Costa Cabral, DC, MSc, ~o Paulo School of Medicine, Sa ~o Paulo, Brazil. E-mail: University of Sa [email protected] 1526-5900/$36.00 ª 2014 by the American Pain Society http://dx.doi.org/10.1016/j.jpain.2014.07.001
as a whole.64 Musculoskeletal disorders have recently been listed as the first cause of ‘‘years lived with disability’’ worldwide49 and are one of the most common reasons for which people seek medical care.12 Estimates of the prevalence of CP range from 10.1 to 55.2%, and a weighted mean prevalence of 30.3% has been suggested.24,31 Cultural aspects and methodologic differences are commonly mentioned as explanations for the wide variation observed across studies.31,33,67 It has been suggested that CP should be considered a distinct diagnostic entity and that genetic, psychological, and social factors contribute to the perception and expression of the condition.59,63 Health measurement instruments are an efficient tool for the standardized identification of attributes of pain, including intensity, disability, persistence, and associated behavioral factors.29,62,69,70 Additionally, the use of previously validated scales may provide a common ground for comparison between studies. 1081
1082
The Journal of Pain
Several factors have been associated with CP. The condition is most commonly reported by females and older persons.6,8,25,33,55,57,67,68 The associated social conditions include unemployment, a lower socioeconomic level, a lower educational level, and employment that requires high physical strain.4,6,8,25,33 Depression and smoking are 2 psychosocial characteristics mentioned.2,4,11,15,21,51,57 Other associated factors include obesity, marital status (divorced or separated), and poor self-rated health.8,25 The current epidemiologic survey on the prevalence of CP in the most populated city in Latin America may provide useful data for the development of effective health care policies in a region where few populationbased studies on this subject have been published.57,68 The discrimination of specific groups according to the severity of symptoms might aid in targeting interventions to persons at higher risk. The present study was conducted to estimate the prevalence of CP in a ~o segment of the general population of the city of Sa Paulo and to identify sociodemographic, psychosocial, and occupational factors associated with CP.
Methods Sampling and Procedures A cross-sectional, population-based epidemiologic ~o Paulo City study was conducted in a central part of Sa under the administration of the Brazilian Family Health Program. All of the families that resided in this area were registered at a specific Health Care Unit in the district of Barra Funda. The community comprised 8,052 individuals from 2,549 families. The households were divided into 17 microareas, each comprising a geographically delimited area with approximately 150 families. A community health agent was responsible for visiting the families from each microarea once per month. Population changes noted during those visits were used to update the records, thus providing a reasonably reliable registry. The current study was approved by the ethics and research committees of Faculdade de Medicina da Universidade de ~o Paulo, Faculdade de Medicina da Santa Casa de MiserSa rdia de Sa ~o Paulo, and the Sa ~o Paulo City Prefecture. ico Participants provided informed consent. Permission was also obtained for the use of the EuroQol-5D. A CP prevalence of 25% was expected based on a ~o recent study that was conducted in the city of Sa Paulo.10 A sample size of 820 participants was estimated, considering a sampling error of 3% and a confidence interval of 95%. The participants were selected via random probability sampling. The number of families selected from each microarea was determined in proportion to the relative size of the microarea. Then, the families, identified by numbers, were randomly selected (Excel Sample Test, 2010.11; Microsoft Inc, Redmond, WA). One individual from each family was chosen from a preordained table of family members listed according to age and gender in relation to the head of the household.34 The inclusion criteria were persons of either gender aged 15 or older. Persons who were unable to participate in the interview
Chronic Pain Prevalence in S~ ao Paulo City for any reason and those who were considered to be dangerous to the interviewers were excluded. Each household was visited at least 4 times at different hours and on different days of the week in an attempt to maximize the chance of locating the person. If contact was not established by the fourth attempt, a new individual from a different family was selected using the same method.
Instruments and Data Collection The data were collected through domiciliary interviews that were conducted by previously trained university students. A pilot study with 30 subjects was conducted at a university ambulatory clinic to train the interviewers and identify the subjects’ perceived difficulties in response to the questionnaire. The interviews consisted of an initial questionnaire, followed by the application of 4 health measurement scales and 1 socioeconomic classification instrument. To gather reliable standardized information on factors associated with CP, validated instruments were used. CP was measured using the Chronic Pain Grade (CPG), an 8-item questionnaire that evaluates the most bothersome pain in the preceding 3 months.69 This instrument classifies CP into 4 hierarchical and mutually exclusive grades. The pain intensity is calculated as the average of 3 questions concerning the present, worst, and average pain, as measured on a 0 to 10 scale, ranging from ‘‘no pain’’ to ‘‘pain as bad as could be.’’ ‘‘Disability days’’ refers to the number of days for which pain prevented usual activities. The pain-related disability is calculated as the average of 3 questions on pain interference with daily activities, social and family activities, and work, as measured on a 0 to 10 scale, ranging from ‘‘no interference’’ to ‘‘unable to carry on any activities.’’ The pain grade is assigned to the following 4 hierarchical and mutually exclusive classes: I, low intensity; II, high intensity; III, moderately limiting; and IV, severely limiting. The psychometric properties of the CPG have been studied in different population settings,9,23,69 and the CGP has been used in epidemiologic surveys,23,41,54,74 clinical trials,66 and studies on risk factors for chronic pain.15,44 Health measurement scales with variables that are known to be associated with CP were also included in the present study. All instruments were previously culturally adapted to Brazilian Portuguese, and validation studies were performed on the Brazilian population. The Hospital Anxiety and Depression Scale (HADS) was developed to identify symptoms that are associated with anxiety and depression in nonpsychiatric individuals.7,16,75 This scale is divided into anxiety and depression subscales, each consisting of 7 intermingled multiple-choice questions. Four alternatives, weighed 0 to 3, are provided for each question. A score above 8 is considered the optimal threshold for anxiety and depression, with a sensitivity and specificity of approximately .80 for both subscales. The HADS has been extensively used and is frequently applied in population studies. Furthermore, its psychometric properties have been studied in several populations.7,54 In the current study, anxiety and depression were considered to be
Maia Costa Cabral et al present with a score equal to or greater than 9 points on the respective subscale. The Alcohol Use Disorders Identification Test was developed by the World Health Organization as a screening instrument for excessive alcohol use. It consists of 10 questions that evaluate the recent use of alcohol, symptoms of dependence, and problems related to alcohol use. Respondents are classified into 1 of the following 3 domains: hazardous alcohol use, dependence symptoms, and harmful alcohol use.39,42 € m Test for Nicotine Dependence is a The Fagerstro screening tool that provides a self-reported measure of nicotine dependence. It consists of an 8-item scale that provides 5 grades of dependence, ranging from very low to very high.14,43 The EuroQol-5D was developed by the EuroQol group as an instrument for the self-assessment of healthrelated quality of life.19,26 It consists of 5 multiplechoice questions that provide scores on the following 5 dimensions: mobility, self-care, usual activities, pain/ discomfort, and anxiety/depression. The final score, known as the EuroIndex, ranges from 0 (the lowest score and worst health status) to 1 (the highest score and best health status). Additionally, the global rating of current health, as assessed by a numeric scale that ranges from 0 (the worse health state) to 100 (the best health state), is commonly used as an indicator of self-rated health. The psychometric properties of the EuroQol-5D have been studied in several populations.38,40 The Brazil Economic Classification Criteria classifies the population into 5 economic segments, ranging from A to E, according to family income.20 During the interviews, the demographic characteristics, pain symptoms, number of days of pain medication use during the past 30 days, and number of pain-related medical visits during the previous 3 months were recorded. CP was considered present when participants reported pain of any intensity for 3 months or more and at least 1 episode of pain during the past 30 days. Persons with CP were asked to draw their pain on a body map of the human figure and to signal with an arrow the site of the most bothersome pain. CPG was then completed based on the location of that pain. Chronic widespread pain was defined according to the criteria of the American College of Rheumatology as pain that is present below and above the waistline and on both sides of the body.73 All participants completed the HADS, Alcohol Use Disorders Identification Test, and Brazil Economic Classification Criteria questionnaires; those who smoked also re€ m Test for Nicotine Dependence. sponded to the Fagerstro
Statistical Analysis Statistical analysis was performed using Stata, 12.0 (StataCorp LP, College Station, TX). Descriptive analyses of the median, mean, standard deviation, and percentage were used to establish the demographic and clinical characteristics of the sample. The CP prevalence and its 95% confidence interval (CI) were determined. The association between CP and the selected variables was estimated by prevalence ratios and 95% CI. Initial
The Journal of Pain
1083
analysis included the chi-square test. For ordered (categorical) variables, the chi-square test for a linear trend in the prevalence ratio was used by categorizing the exposure variables and entering the continuous scores. Variables with a P value less than .20 in univariate analysis were selected for a multiple regression analysis. We used a Cox regression model with constant time and robust variance.5,36,37 Usually, Cox regression is employed to analyze time-to-event data. When a constant risk period (time = 1) is assigned to all subjects, the hazard rate ratio estimated by Cox regression equals the prevalence ratio in cross-sectional studies.5,36 The final model included only those variables with P
