Systematic Review of Multidisciplinary Chronic Pain Treatment Facilities

This study reviewed the published literature evaluating multidisciplinary chronic pain treatment facilities to provide an overview of their availability, caseload, wait times, and facility characteristics. A systematic literature review was conducted using PRISMA guidelines following a search of MEDLINE, PsycINFO, and CINAHL databases. Inclusion criteria stipulated that studies be original research, survey more than one pain treatment facility directly, and describe a range of available treatments. Fourteen articles satisfied inclusion criteria. Results showed little consistency in the research design used to describe pain treatment facilities. Availability of pain treatment facilities was scarce and the reported caseloads and wait times were generally high. A wide range of medical, physical, and psychological pain treatments were available. Most studies reported findings on the percentage of practitioners in different health care professions employed. Future studies should consider using more comprehensive search strategies to survey facilities, improving clarity on what is considered to be a pain treatment facility, and reporting on a consistent set of variables to provide a clear summary of the status of pain treatment facilities. This review highlights important information for policymakers on the scope, demand, and accessibility of pain treatment facilities.


Introduction
Chronic pain is a major global health problem that affects approximately 37% of individuals in developed countries [1]. In Canada, it is estimated that as many as 25% of adults have a chronic pain condition [2]. Chronic pain places a substantial burden on the individual and can contribute to disability [3], depression [4], anxiety [5], and lowered quality of life [6]. There are also significant societal costs associated with chronic pain. In the United States, chronic pain is estimated to cost between 560 and 635 billion dollars annually [7], while, in Canada, direct costs to the health care system are estimated to be over 6 billion dollars and indirect costs due to job loss and sick days are estimated to be over 37 billion dollars annually [8,9].
The development and maintenance of chronic pain is complex, affected by biological, psychological, and social factors [5,10]. Consequently, the most effective pain treatment strategies target a variety of factors simultaneously, an approach that is used in multidisciplinary pain treatment facilities [11][12][13]. The services and standards of multidisciplinary treatment vary considerably, which led the International Association for the Study of Pain (IASP) to release guidelines regarding the classification and ideal standards of pain services [14]. IASP defines a "multidisciplinary pain center" (a Level 1 facility) as being staffed by a variety of health care professions with expertise in pain management, including physicians, nurses, mental health professionals, and physical therapists. The team should work together and be able to effectively assess and treat any pain problem. Training opportunities, education, and research should be available at the pain center. A "multidisciplinary pain clinic" (a Level 2 facility) must uphold the same standards of a multidisciplinary pain center, although they may not have training and research opportunities. Further classification includes "pain syndrome programs" that provide multidisciplinary care for one type of pain and "single modality therapy programs" that provide one type of treatment. A "pain treatment facility" is a general term that encompasses all of the above definitions [15].

Pain Research and Management
Despite being considered the highest standard of care for chronic pain, there is limited research regarding the availability and characteristics of pain treatment facilities globally. In order to provide a comprehensive overview of the published literature, the Canadian Pain Coalition (http://www.canadianpaincoalition.ca/) sponsored the present systematic review as the first phase of a "Report Card on Pain." This initiative aims to identify areas of strength and weakness in pain treatment facilities worldwide to help inform policymakers on targeted funding strategies to better tackle the challenge of pain in Canada and globally. The objective of the present paper is to systematically review the published literature of surveys on multidisciplinary pain treatment facilities to provide an overview of their availability, caseload, and wait times, as well as the facility characteristics, including available treatments and employed pain professionals.

Search Strategy.
The systematic review was conducted according to PRISMA guidelines [16]. The review protocol was developed by the authors prior to beginning the searches. Eligibility criteria required articles to meet four study characteristics: (1) studies must describe pain treatment facilities directly with questionnaires or other data collection methods (excluding qualitative studies); (2) studies must be original research articles (i.e., no editorials, reviews); (3) more than one multidisciplinary pain treatment facility must be surveyed; and (4) questionnaires must describe a range of pain treatments rather than focusing on one type or category of treatment (e.g., acupuncture). At this stage, no restrictions were placed on report characteristics (e.g., language), publication status, location of study, or date of publication. Additionally, studies were not required to be published by an academic journal and could be published by an established organization.
The search strategy included searching academic article databases, asking experts in the field, and reviewing reference lists of included articles. No language restrictions were made for the initial search. Database searches were conducted using MEDLINE (1946 to August 2014), PsycINFO (1912 to August 2014), and CINAHL (1979 to August 2014). The last search was completed on September 23, 2014. All authors developed the search protocol and one author (Joel Katz) conducted the searches. The authors of all included articles were contacted for the original questionnaire used in their study.
The following search terms were used: pain; pain clinic * ; pain service * ; pain cent * ; pain facilit * ; pain management clinic * ; pain management service * ; pain management cent * ; pain management facilit * ; pain treatment clinic * ; pain treatment service * ; pain treatment cent * ; pain treatment facilit * ; survey * ; program evaluation; program assessment; program outcome; questionnaire * . See Appendices A and B for a complete description of the search strategy and search terms.

Data Collection.
Study titles and abstracts were reviewed for inclusion by one author (Joel Katz) according to the eligibility criteria. Articles selected for full-text review were independently reviewed by two authors (Joel Katz and Samantha R. Fashler). Disagreements were discussed until a consensus could be reached.
Data was transposed into a data extraction spreadsheet collaboratively created by three authors. The spreadsheet was modified as needed to accommodate new information from studies as they were reviewed. One author entered the extracted data from the studies into the spreadsheet (Samantha R. Fashler) and one author checked the data that was entered (Eric D. Oosenbrug). Study authors were contacted to receive a copy of the survey used in the project. Six authors provided a copy of the survey, one author responded to say they did not have a copy of the survey, five authors did not respond, and one survey was available in the appendix of the published study. In one case, results on the same survey were reported in two studies [17,18]. To avoid bias, no duplicate information was extracted and nonoverlapping content was included in the extraction spreadsheet.
All variables included in the data extraction spreadsheet related to the assessment and evaluation of pain treatment facilities. Specifically, information related to following domains was extracted: (1) study characteristics, including a definition of pain treatment facilities used, search strategy, location and date surveyed, questionnaire used, and the response rate; (2) access to pain treatment facilities, including the rate of new and follow-up consultations and wait times; (3) pain treatments available, including medical approaches, physical therapies, and psychological services; and (4) characteristics of pain treatment facilities, including the employed health care professionals, staff, space, and equipment.
Risk of bias was assessed with a tool developed by the Cochrane Collaboration [19]. It is a domain-based evaluation tool, assessing the following: selection bias, performance bias, detection bias, attrition bias, reporting bias, and other biases (not addressed with the former biases).
Summary measures included standardized differences in means. Data transformation was used to increase comparability across reported data: specific alterations to original data are described in their respective table. No additional analyses were undertaken.

Study Selection.
Searches conducted in MEDLINE, PsycINFO, and CINAHL yielded 2117 hits. Eighteen additional articles were hand selected by consulting with experts and searching reference lists. After duplicates were removed, 1692 unique articles remained. Articles were first screened by title and abstract by one author (Joel Katz), eliminating 1627 articles. The full texts of the remaining 65 articles were reviewed by two authors (Joel Katz and Samantha R. Fashler). Fifty-one articles were eliminated because they did not investigate pain treatment facilities ( = 26) or because they did not meet the selection criteria: eight articles surveyed clinicians or patients directly rather than pain treatment facilities, three only looked at one pain treatment facility, and three articles only examined one type of service.  study design aimed to include non-English articles. However, the 11 non-English articles identified were eliminated from the present review due to difficulties in accurately translating them. See Figure 1 for a flow diagram of the articles selection and inclusion process used. In total, 14 studies were included in the present review. Reporting bias may be present in the studies included in the review. When the questionnaires were available, the reported outcomes in the paper were compared to the questionnaire used to determine if any relevant data were not reported. The studies included in the review had a recruitment bias, since many pain treatment facilities that were contacted did not respond to the surveys, and are therefore not represented in the reported survey findings. A selection bias was present for the baseline characteristics of the surveys as the definition of a pain treatment facility was different for each study.

Study Characteristics.
The 14 included studies were published between 1985 and 2013 [17,18,[20][21][22][23][24][25][26][27][28][29][30][31]. Surveys were carried out around the world, with five in the United Kingdom, four in Canada, three in the United States, one in Australia, and one in Italy. Surveys reported countrywide or region-specific data. Ten were published in scientific journals [17,18,[20][21][22][23][24][29][30][31], two were audits of pain services [27,28], and two were reports published by professional organizations [25,26]. Despite examining the same topic, each study used different search criteria to find pain treatment facilities and used different eligibility criteria for what was considered a pain treatment facility (see Table 1). The response rate in studies ranged from 56% to 100% (M = 80.1%; 95% CI: 70.92 to 89.31). Each study has a bias according to the time it was conducted, location of the survey, the manner in which it was published (i.e., in a peer-reviewed journal or an independently published report or project), pain treatment facility search strategy, eligibility criteria, and response rate.
All studies used a survey that was sent by phone, fax, email, or mail. Only one study [20] included an interview as a part of data collection. Pain treatment facility directors, chief executive officers, head anesthesiologists, and, in one case, senior psychologists were asked to complete the questionnaire. Surveys were developed for each study with two exceptions: Peng et al. [17,18,21] used an adapted version of the Québec Chronic Pain Clinic Survey developed by Veillette et al. [22] and Csordas and Clark [30] did not specify what survey was used.
Health care systems can vary substantially depending on the country of origin [32]. For this reason, the results on pain facility characteristics are reported by country (Australia, Canada, Italy, United Kingdom, and United States).

Availability and Caseloads of Pain Treatment Facilities.
Data on the availability, caseloads, and wait times for pain treatment facilities are listed in Tables 2 and 3. In many (1) Space and beds exclusively available to the pain service   Note. Information on incidence and caseload was not available for Bisset [24], Csordas and Clark [30], National Pain Audit [28], and Peng et al. [17]. a Divided by 12 to provide a monthly estimate since annual appointments were reported in source paper. b Divided by the number of pain treatment services surveyed since the total number of appointments was reported in the source paper. c Multiplied by 4.35 to provide a monthly estimate since weekly appointments were reported in source paper.
cases, the metric used to report this information varied, with some studies reporting medians, means, ranges, or sums. To facilitate meaningful comparisons across studies, relevant data were transformed to match the metrics reported in other studies (e.g., if new consultation appointments were reported annually, it was divided by 12 to permit comparison to other articles that reported this index according to new consultations per month).
Australia. In their survey, Hogg et al. [20] found that there was only one pain treatment facility for 310,000 Australians.   Tables 4-6).
Australia. Only one form of pain treatment was reported by Hogg et al. [20]. They reported that 3.5% of pain treatment facilities used acupuncture.
Canada. In three Canadian surveys [18,21,22], medical treatments included cryotherapy (6%), injections [trigger point injection (60-88%), botulinic toxin injection (26-44%), continuous epidural (80%), one-shot epidural or epidural injection (46-100%), facet joint/nerve injection (51-60%), and intravenous regional anaesthesia (   Note. Information on medical treatments was not available for National Pain Audit [27], [28] and Peng et al. [17]. Hogg et al. [20] did not state specific procedures but indicated that 87.7% of pain treatment facilities used some sort of interventional medicine procedure. De Benedittis and Lorenzetti [23] reported a mean utilization index instead of the percentage of facilities that used a specific procedure and data were therefore not included in the table. Articles reporting on a procedure not reported in another study were not included in the table ( = 27 procedures). Articles reporting treatments without percentages were not included.

86.7%
Note. Information on physical therapy related treatments was not available for National Pain Audit [27], [28], Peng et al. [17], and Veillette et al. [22]. De Benedittis and Lorenzetti [23] reported a mean utilization index instead of the percentage of pain treatment facilities that used a specific procedure and data were therefore not included in the table. Articles reporting on a procedure not reported in another study were not included in the table ( = 11 procedures). Articles reporting treatments without percentages were not included.
Information on psychological and other treatments was not available for Hogg et al. [20], National Pain Audit [27], [28], Peng et al. [17], and Veillette et al. [22]. De Benedittis and Lorenzetti [23] reported a mean utilization index instead of the percentage of pain treatment facilities that used a specific procedure and data were therefore not included in the table. Articles reporting on a procedure not reported in another study were not included in the table ( = 17 procedures). Articles reporting treatments without percentages were not included. a Reported as family/couple therapy.
Italy. De Benedittis and Lorenzetti [23] reported a mean utilization index instead of the percentage of facilities that used specific treatments and data was therefore not included in the tables. Of treatments that were reported in at least one other study, they reported that pain treatment facilities used nerve blocks, percutaneous discectomies, pharmacotherapy (including psychotropic medication, long-acting narcotics/opioids, and other medications), and spinal cord stimulation.
United Kingdom. Three surveys in the United Kingdom reported on treatments used in pain treatment facilities [24][25][26].   Note. Information on employed pain professionals was not available for Dr. Foster & the Pain Society [26], Hogg et al. [20], National Pain Audit [28], and Peng et al. [18]. Articles reporting on a profession not reported in another study were not included in the table ( = 29 professions). Articles reporting professions without a percentage were not included.
a Reported only for services with specialist medication management.
b Listed in the paper as either a physical or occupational therapist.
c Listed in the paper as either psychiatrists or psychologists.  Note. Information on multidisciplinary status was not available for Bisset [24], Csordas and Clark [30], and Veillette et al. [22]. IASP: International Association for the Study of Pain. PMP: pain management program.
a Training for medical and physiotherapy students. b Training for nurses.
c Training for medical students.

Pain Research and Management
Australia. Information on pain professionals working at pain treatment facilities was not reported by Hogg et al. [20], although information on space and multidisciplinary status was available. It was reported that 60% of pain treatment facilities had inpatient care/beds [20]. The authors reported that 46% of pain facilities met Level 1 IASP criteria [14] (i.e., a multidisciplinary pain center) and 33% met Level 2 criteria (i.e., a multidisciplinary pain clinic). Additionally, 33.3% of facilities offered training and 74% offered a pain management program (PMP) [20].

Discussion
The present study systematically reviewed the literature on surveys of pain treatment facilities. Our search yielded 14 English language studies published between 1985 and 2013. Surveys were conducted in Australia, Canada, Italy, the United Kingdom, and the United States. The results reveal little consistency in the research design and questionnaires used to describe pain treatment facilities. Each study used different search and inclusion criteria for pain treatment facilities. The availability of pain treatment facilities was scarce even though the caseloads and wait times for facilities were generally high. A wide range of interventional medical, physical, and psychological pain treatments were available.
Most studies reported findings on the percentage of practitioners in different health care professions employed. Taken together, these findings suggest that there is considerable variability in the availability, services, and professionals of multidisciplinary pain clinics. Each survey used a different approach to identify pain treatment facilities (Table 1). In some cases, facilities were identified through contacting individuals with membership in a specific association (e.g., [24]), searching the Yellow Pages (e.g., [29]), consulting a published list of facilities (e.g., [25,31]), or using a combination of different strategies. When the search for pain treatment facilities is not comprehensive, there is a risk of underestimating the availability of pain treatment facilities in a region and of selection bias that may undermine the representativeness of sampled pain facility characteristics. For example, Veillette et al. [22] identified pain treatment facilities by contacting the heads of anesthesia departments, a strategy that may have omitted facilities that are not directed by individuals in that role. This suggests that caution must be taken when interpreting and generalizing findings.
Pain treatment facilities are both scarce and in high demand (Tables 2 and 3). Access to multidisciplinary care varied by country, with one per 310,000 people in Australia [20], one per 258,000 people in Canada [18], and one per 200,000-370,370 people in the United Kingdom [27]. With an estimated 37% of the world population suffering from chronic pain [1], this reflects poor availability of services for pain sufferers. The demand for multidisciplinary pain treatment facilities is reflected in the wait times.
In 2005, IASP [33] developed recommendations for wait times based on an international task force of experts. They suggest wait times for acute painful conditions to be immediate, most urgent pain to be within one week, urgent or semiurgent pain to be within one month, and routine or regular pain to be within eight weeks. In the present study, we found one article from the United Kingdom that reported on wait times according to the severity of painful conditions [25] that showed the median wait times to be two weeks for urgent care and 16 weeks for routine care, both substantially longer than the suggested wait times. Two surveys showed breakdown of wait times according to public and private facilities [18,20]. It was found that wait times were longer in public facilities in Australia (Mdn = 21.4 weeks) and Canada (Mdn = 25.7) in comparison to private facilities (Australia: Mdn = 5.5, Canada: Mdn = 2.1), showing that those that can afford to pay out of pocket for medical services can receive multidisciplinary pain treatment faster. No information was available on wait times from surveys conducted in the United States, where public funding is mainly limited to individuals that are 65 years old and older, individuals with disabilities, and some individuals with low income, although there are substantial gaps in health care coverage for the general American population [34]. It is possible that, in the United States, there may be increased demand for private services that may increase wait times. Importantly, it appears that access to the multidisciplinary care may be faster in pediatric settings: in a survey of five pediatric pain treatment facilities, Peng et al. [21] reported the median wait time to be four weeks, although it was not specified whether these facilities were public or private. Taken together, as longer wait times are associated with deterioration of health and lowered quality of life [35], the present findings suggest an alarming global situation in which access to pain treatment facilities requires improvement.
Unsurprisingly, it was found that medical pain treatments were reported more frequently in pain treatment facilities in acute care hospitals in Québec [22] in comparison to pain treatment facilities not restricted to hospital settings that were surveyed across Canada [18,21]. The emphasis on medical treatments by Veillette et al. [22] is also illustrated in the exclusion of data collected on the prevalence of physical therapy and psychological and other treatments that were reported in Canada-wide surveys [18,21]. Surveys conducted in the United Kingdom were published between 1988 and 2003. Despite the range, the prevalence of some reported treatments did not vary greatly. For example, continuous epidural was reportedly used in 68.8% of facilities in Bisset [24], 73% in Clinical Standards Advisory Group [25], and 61% in the report by Dr. Foster & the Pain Society [26]. However, it is noteworthy that only treatments reported in at least two studies (regardless of country of origin) were included in the review. In the three surveys conducted in the United States, the prevalence of injections and nerve blocks was only reported in two studies, with more information reported on pharmacotherapy treatments. Little information was available on pain treatments for surveys conducted in Australia and Italy, limiting the discussion on their use of interventions in pain treatment facilities.
According to IASP, multidisciplinary pain centers and clinics must be staffed by a variety of health care professionals that include physicians, nurses, mental health professionals, and physical therapists [14]. All Canadian surveys reported on the prevalence of these employed professionals, even though they were not always represented. However, in pediatric settings, Peng et al. [21] found that 100% of pain treatment facilities staffed a psychologist and a nurse, while 80% staffed an anaesthesiologist and physiotherapist. In Italy, the majority of pain facilities surveyed by De Benedittis and Lorenzetti [23] did not satisfy the staff requirements of a multidisciplinary pain centre or clinic given that physical therapists were reported in only 14.3% of facilities and mental health professionals were reported in 30.2%. Data on the staff composition from surveys in the United Kingdom did not include any information on employed physicians, although the authors reported that nurses were staffed in up to 93.8% of facilities [24], physiotherapists in up to 100% [24], and psychologists in up to 67% [25]. In surveys conducted in the United States, Castel et al. [29] found that only a minority of facilities met IASP criteria for a multidisciplinary clinic and this low number was attributed to only having one provider type on staff. This may be related to the methodology used in this survey, as only pain facilities that specialized in treating neck and back pain were contacted for the survey. Although the staff composition was not surveyed by Hogg et al. [20], they reported that 46% of pain treatment facilities in Australia met IASP criteria for a multidisciplinary pain centre. Overall, even though many studies did not include a measure of the multidisciplinary functionality, an assessment of facilities provided demonstrates that pain treatment facilities attempt to provide comprehensive treatment to target both biomedical and psychological dimensions of pain. This study has several limitations. The data extracted from the studies included in the review were heterogeneous, with surveys completed over a 28-year period and in five countries. There are considerable differences in what is considered to be a pain treatment facility and the importance of various treatment strategies has changed over time [14,36]. The method of identifying pain treatment facilities and the differing eligibility criteria introduced bias where not all multidisciplinary pain clinics may have been included in the reported surveys. This may have been similarly affected by the responses rate, which ranged between 56 and 100%. Additionally, drawing inferences across the extracted data sample was challenging due to the extent of missing data and differently reported metrics.
Regular, comprehensive surveys and audits of pain treatment facilities are warranted. They provide important information on the availability of services, wait times to access treatment, and the types of services available. There were also many variables of interest for future research that were assessed in some studies but not others (see Table 9). These include information regarding the cost of treatments, outcome measures used, whether facilities are public or private, and guidelines that are currently in use. Future investigations should consider these variables. The Canadian Pain Coalition conducted this study as a part of the first phase of the Report Card on Pain. For the second phase, the Report Card on Pain aims to complete an updated survey of pain treatment facilities in Canada, as the most recent survey completed was over 8 years ago [18].
In summary, this systematic review of studies surveying pain treatment facilities around the world highlights important information for policymakers on the scope, demand, and accessibility of such facilities and provides direction for future researchers to enhance the comprehensiveness and comparability of data capture. The results showed that there is a great deal of diversity in the information collected and future studies should consider using more comprehensive search strategies, increasing clarity on what is considered to be a multidisciplinary pain treatment facility, and reporting on a consistent set of variables to provide a clear summary of the status of multidisciplinary pain treatment facilities. Despite their variability, the identified studies shed light on substantial gaps in tackling the challenge of pain and underscore the pressing need for improved access to pain treatment facilities globally.

Summary
This study systematically reviewed the published literature describing multidisciplinary chronic pain treatment facilities to provide an overview of their availability and characteristics. Fourteen articles satisfied inclusion criteria (i.e., original research that directly surveyed a range of treatments in more than one pain treatment facility). Results show considerable variability in the research methodologies used to describe pain treatment facilities, show that availability of pain treatment facilities was scarce, and show that accompanying caseloads and wait times were generally high. Pain treatment facilities used a wide range of pain interventions and employed a variety of health care professionals. Although the present study was limited by the heterogeneous data extracted from included studies, it highlights important information on the scope, demand, and accessibility of multidisciplinary pain treatment facilities.