|The London Fibromyalgia Epidemiology Study (LFES):
The Prevalence and Direct Health Care Costs of fibromyalgia syndrome (FMS)
in London, Ontario
Kevin P. White
The University of Western Ontario,
Background: Major objectives of the current study were to 1) estimate the point prevalence of fibromyalgia syndrome (FMS) among non-institutionalized Canadian adults, and 2) to estimate the direct health care costs associated with FMS.
Methods: This study was conducted in three phases. In Phase I, a validated screening questionnaire was administered via telephone to a random community sample of 3395 non-institutionalized adults residing in London, Ontario. Individuals screening positive were invited into Phase II, during which one of two rheumatologists confirmed or excluded FMS using the 1990 American College of Rheumatology (ACR) classification criteria. In Phase III, direct health care costs were compared among those with confirmed FMS, those with widespread pain not meeting the ACR criteria, and a control group not reporting widespread pain.
Results: FMS affects an estimated 4.9% (95 % confidence intervals = 4.8%, 4.9%) of adult females and 1.6% (1.5%, 1.8%) of adult males in London, Ontario. In women, the prevalence of FMS rises steadily from less than one percent in the age group 18-30 years, to a peak of almost eight percent in the age group 55-64. Then, it declines. The peak prevalence of FMS in males also appears to be in middle age (2.5%; 1.6%, 4.0%). No cases of FMS were confirmed in males over age 64 years. FMS affects 3.3% (3.2%, 3.4%) of the non-institutionalized adult population, approximately 700,000 Canadian adults. Less education, lower household income, being divorced, and being disabled are associated with increased odds of having FMS. Individuals with FMS use more medications and outpatient health services than those with chronic widespread pain without FMS, and approximately twice the health services at twice the cost as the general population. The mean difference between community cases of FMS and matched controls from the general population, with respect to direct costs for health services, is $493 annually (p < 0.001). This extrapolates to annual estimated direct health care costs attributable to FMS in Canada between $300 and $350 million.
Conclusions: FMS is a common musculoskeletal disorder, especially in females, that has a major impact on Canadian health care costs. That its prevalence declines in females over age 64 years cannot easily be explained by the traditional view that it is a chronic, non-remitting, non-crippling and non-fatal disorder.
Musculoskeletal (MSK) disorders contribute greatly to chronic disability and health services utilization in developed economies. In the USA, an estimated 37 million individuals (14.5% of the U.S. population) suffered from arthritis in 19891. MSK disorders account for 15% of work loss days in the USA2 and 14 -17% of work loss days in Great Britain3. In Canada, data suggest that one million adults have physical disabilities secondary to MSK illness, a prevalence of 50.1 per thousand4. Sixteen percent of respondents in the 1980 Canadian Health Survey (CHS)5 and 14 percent in the 1990 Ontario Health Survey (OHS)6 reported having either arthritis or rheumatism. In each study, ‘arthritis and rheumatism’ was the most commonly reported cause of chronic disability, and the second most frequent cause of two week disability (after respiratory illness including rhinitis). These findings were consistent across all age groups. In addition, people affected by arthritis and rheumatism use more health services than the general population7, and the costs of MSK illness may be increasing. In 1980, MSK disorders in the U.S. accounted for an estimated $21 billion in health care costs and lost wages, approximately one percent of the country’s Gross National Product8. A more recent survey, performed between 1990 and 1992, estimated the same costs as $149.4 billion, or 2.5% of the GNP9.
Fibromyalgia Syndrome (FMS), also known as fibrositis, is a common form of non-articular rheumatism associated with chronic generalized musculoskeletal pain, fatigue, and a long list of other complaints10. Clinic studies indicate that FMS patients are as adversely affected by their symptoms as patients with rheumatoid arthritis (RA), and more so than osteoarthritis (OA) patients11,, For several reasons, including a relative dearth of specific physical, laboratory and radiographic findings, the concept of FMS has met with much skepticism.,,, and has attracted less research interest than other less common musculoskeletal disorders12. Despite this, it is diagnosed in 10 to 20 percent of patients seen in outpatient rheumatology clinics in the U.S.13, Mexico14, Spain15, and Australia16 and appears to be one of the three most common disorders diagnosed among new referrals to Canadian rheumatologists17. The prevalence of FMS in the adult U.S. population has been estimated at 2.0%, based upon 36 cases of FMS that were confirmed in a general population study in Wichita, Kansas18.
To date, the prevalence of FMS in Canada is unknown. The first objective of this study was to estimate the point prevalence of FMS among non-institutionalized adults, in a mid-size Canadian city. The second objective was to identify between 50 and 100 confirmed community cases of FMS, in order to estimate direct health care costs for individuals with FMS, compared to several general population control groups. Previous population-based samples from other countries generally have been too small to assess clinical characteristics and costs of FMS. We will present data on 100 confirmed community cases of FMS. To our knowledge, this represents the largest cohort of community FMS cases reported. These cases will be compared to 1) 76 general population controls with chronic widespread musculoskeletal pain in whom FMS has been excluded; 2) 135 group matched general population controls without chronic widespread pain; and 3) 384 paired matched general population controls from the Ontario Health Insurance Plan (OHIP) billing database.
This was a three phase study. Phase I was a random telephone survey of non-institutionalized adults to identify possible cases of FMS. In Phase II, individuals with widespread, chronic pain were examined to identify FMS cases and collect additional health data. In Phase III, we compared direct health care costs in FMS cases versus controls. The protocol was approved by the Review Board for Research involving Human Subjects at the University of Western Ontario, London, Ontario.
Phase I: Screening for chronic widespread pain.
The target population was London, Ontario, a community of 341,320 persons (1991 census) in Southwestern Ontario. London was selected because of its relatively homogeneous population with respect to race and language, and because it is demographically representative of other mid-size Canadian cities outside of Quebec.
The sampling frame was a computer-generated list of 19,500 seven digit telephone numbers, utilizing an equal number of telephone numbers from each of the 39 telephone exchanges in London. Each sampling unit was a London residence with at least one telephone number, selected by random digit dialing (RDD); approximately 1% of London households do not have telephone service. One person per household was eligible to be interviewed, that being the adult (age 18 years plus) with the most recent birthday as of the date of initial telephone contact; this method for subject selection has been validated previously in the survey literature19.
Required sample size was calculated based on two primary objectives: 1) to estimate overall FMS prevalence with a precision of plus or minus 1.0% with 95% confidence (assuming that the true prevalence of FMS in the study population is 2.0%, and that both a 75% Phase I participation rate and a 75% Phase II participation rate could be achieved); and 2) to identify at least 50 confirmed community cases of FMS. The decision to confirm 50 cases of FMS was made to allow an estimate of mean age with a precision of plus or minus 3 years with 95% confidence, and also to allow for 80% power to detect a $200 or greater difference in annual OHIP spending between study groups. To meet these objectives, we estimated that 4500 eligible subjects should be contacted in the telephone survey, resulting in 3375 completed screening interviews.
Subjects screened positive if they reported widespread pain of at least one week’s duration over the preceding three months. In addition, subjects who screened negative, but reported having been told that they had fibrositis or fibromyalgia, were re-contacted by a rheumatologist within one week of their initial interview, to confirm their earlier responses to the pain items. This was done to reduce the risk that previously diagnosed FMS cases who misunderstood the screening questions would be missed. A subject who screened positive upon re-testing within one week would be invited into Phase II (no such subjects were identified).
Also during the telephone interview, subjects were asked their sex, birth year, birth month, marital status, work status, number of months working in past 12 months, full-time or part-time work status, highest education level achieved, and estimated annual household income, in addition to the number of adults residing in their household. All questions on demographics were selected from the Ontario Health Survey (OHS)20 to yield comparable responses.
Phase I data were collected from November 1994 to March 1996, inclusive.
Phase II: Confirming or excluding FMS; data collection.
All subjects who screened positive in Phase I were invited to be evaluated by one of two rheumatologists (KPW, MH) to confirm or exclude FMS. The evaluation consisted of questions on the distribution and duration of pain, followed by digital palpation for tenderness at the eighteen fibromyalgia points specified by the American College of Rheumatology (ACR)21. Inter-rater reliability previously had been tested in a pilot study of 25 consecutive rheumatology clinic outpatients, of whom 12 had FMS; the two examiners agreed on the presence or absence of FMS in all cases ( = 1.00). Based upon the results of the examination, subjects were identified either as FMS cases (FC) or pain controls (PC). FC and PC then completed a detailed health questionnaire, including: 1) 12 items on the use of medications and drugs over the preceding 4 weeks, selected from Section B of the Ontario Health Survey (OHS) questionnaire; and 2) 18 items on the use of health services over the preceding 12 months, and the preceding 14 days, from Section C of the OHS.
For each FC confirmed in Phase II, two general controls (GC) without chronic, generalized pain, matched for age and sex, were selected during subsequent telephone interviews. The same Phase II Health Questionnaire that was completed by all FC and PC subjects was mailed to all GC subjects, accompanied by an introductory letter, an information letter and consent form, and a pre-addressed, pre-stamped envelope. GC subjects who had not returned the completed questionnaire were re-contacted by telephone within three weeks.
Estimating Point Prevalence
Three estimates of point prevalence were calculated. The most conservative estimate was based upon the assumption that the only cases of FMS in the survey sample were those cases confirmed in Phase II.
Estimate #1: confirmed FMS cases x 100%
total Phase I participants
A second estimate adjusts for Phase II non-participation by assuming the same prevalence among positive screens who refused to be examined as among those who participated in Phase II.
Estimate #2: confirmed FMS cases + (p x Phase II refusals) x 100%
total Phase I participants
where p = the proportion of Phase II participants in whom FMS has been confirmed.
A third estimate adjusted both for Phase II non-participation and the number of adults residing in each household22.
Estimate #3: (confirmed FMS cases x ni) + (p x (Phase II refusals x ni) x 100%
(Phase I participants x ni)
where ni = the number of adults living in the ‘i’ household.
Each of the above estimates was adjusted both for age and sex by direct age standardization, using 1991 London census data. Ninety-five percent confidence intervals were constructed using logit transformation to correct for proportions approaching zeroth.
Using subject responses to demographic items in the Phase I telephone interview, we calculated the crude odds of having FMS for each demographic variable, with 95% confidence limits.
Phase III: Estimating Direct Health Care Costs
All FC, PC and GC subjects were asked to participate in a group comparison of annual health services costs to the provincially managed, public health insurance program, the Ontario Health Insurance Program (OHIP). In Ontario, OHIP is the primary source of reimbursement for physician, laboratory and imaging services, and also covers many services provided by non-physician health care providers. The OHIP detailed claims file for each consenting subject was entered to access all individual billing claims for the twelve months of 1994, the year prior to Phase II data collection. For each subject, health care utilization for the year 1994 was calculated as 1) the quantity of health services used in 1994, and 2) the cost, in 1994 Canadian dollars, of health services used in 1994. Claims then were subdivided into: 1) reimbursement of physicians, 2) reimbursement of other health care professionals, 3) laboratory costs, and 4) radiology costs. These costs then were aggregated for the three groups.
Phase III Data Analysis
The SAS statistical package23 was used to analyze group differences in mean annual number of services and annual OHIP costs, using 1) linear regression adjusted for age and sex and 2) where appropriate, the Wilcoxon Rank Sum test to adjust for non-homogeneous group variances. Post-hoc testing, for parametric and non-parametric analyses, respectively, involved: 1) Scheffe’s test to adjust for groups of different sizes, and 2) the Kruskal-Wallis test.
FC also were compared to age, sex and geographically-matched controls from the OHIP database, with four OHIP controls (OC) randomly selected for each FC. Group means for annual number of services and annual OHIP costs were compared using Student’s t-test for paired samples; the mean of each matched group of four OC subjects was calculated and compared to the correspondingly matched FC subject.
To achieve the required sample size, 16,769 telephone numbers were dialed, resulting in 4674 eligible contacts, of whom 1279 (27.4%) refused to participate. This resulted in 3,395 completed screening interviews: 2,090 (61.6%) female, 1,290 (38.0%) male; 15 were unreported and excluded from further analysis (0.4%). Because the sample sex distribution differed from the 1991 London Census (females = 52.7% of adults) all subsequent data analyses were adjusted for age and sex, where appropriate. Among females, the proportion of the total population sampled was least for the 55 to 64 year old age group (1.35%), and greatest for women 35 to 44 (1.72%). Among males, the proportion of the total population sampled was least for men age 55 to 64 (0.76%), and greatest for men 18 to 24 (1.28%).
Pain and fatigue: Recalling the previous three months, 34.8% of the survey sample reported having had some musculoskeletal pain lasting at least one week; this accounted for 36.1% of females, and 32.6% of males (Odds ratio [OR] = 1.16, 95% exact confidence limits [CI] = 1.00, 1.35). A much smaller percentage, 7.3%, reported having had chronic widespread pain, defined as pain lasting more than one week involving the neck, back or chest, plus at least one upper and one lower extremity on opposite sides of the body. Widespread pain also was more commonly reported by females (9.0% versus 4.7%, OR = 2.02, CI = 1.49, 2.77). This 7.3% (248 subjects including 187 females, 60 males, and 1 sex not reported) was invited to participate in the second Phase of the study, as described below.
More than half of the survey sample, 54.5%, reported having had frequent fatigue over the previous three months and, for 21.5%, this fatigue was debilitating enough to significantly limit their activities. As with chronic pain, females were more likely to report fatigue than males, 60.0% versus 45.0% admitting to frequent fatigue (OR = 1.84, CI = 1.60, 2.13), 25.1% versus 15.3% admitting to activity limiting fatigue (OR = 1.86, CI = 1.55, 2.25).
Women also were more likely to report previously having been told they had arthritis (26.3% versus 16.6%, OR = 1.80, CI = 1.50, 2.15) and fibromyalgia (8.6% versus 4.9%, OR = 1.83, CI = 1.35, 2.51).
Phase II participants and non-participants: All 248 subjects who reported having had chronic widespread pain were invited to participate in the second phase of the study; 176 agreed. They were compared to the 72 who refused to participate with respect to demographic characteristics, pain distribution, and degree of fatigue. A somewhat greater percentage of participants were female (79.0% versus 69.4% males) and non-participants on average were older (mean age = 3.2 years); otherwise, the two groups were demographically similar. Phase II participants and non-participants did not differ with respect to the distribution of their chronic pain, nor did they differ with respect to the percentage with frequent fatigue (92.0% versus 91.7%) or fatigue that significantly limited their activities (71.0% versus 63.9%). In females, the phase II participation rate was lowest in the 75 plus age group at 30% [Table 1]. The participation rate in all other age groups was between 66.7% (age 18 to 24) and 80% (age 55 to 64). In males, the phase II participation rate was lowest in the 25 to 34, 45 to 54, and 65 to 74 age groups (50%), and 100% in the 18 to 24 age group [Table 3].
FMS prevalence in females: Among the 184 women who screened positive for chronic widespread pain, 86 cases of FMS were confirmed, and a further 31 cases expected, for a total of 117 cases [Table 1]. Not adjusting for Phase II non-participation, point prevalence is 4.2% (CI = 4.1%, 4.2%) [Table 2]. Adjusting for non-participation, the estimate is 5.7% (CI = 5.5%, 5.7%). Weighting according to the number of adults in each household results in a prevalence estimate of 4.9%. In all three estimates, prevalence starts low at one percent or less, gradually increases until late middle age, then steadily declines [Figure 2].
FMS prevalence in males: Fourteen cases of FMS were identified among 39 male subjects who were examined [Table 3]. The unadjusted, adjusted, and adjusted plus weighted estimates for FMS prevalence in males were 1.0%, 1.7% and 1.6%, respectively [Table 4]. Similar to females, the prevalence of FMS in males is approximately one percent in the 18-24 year age group. Unlike females, the prevalence remains low throughout life, between one and two percent [Figure 2]. There were no cases identified in men over 64 years.
Overall point prevalence: A total of 100 cases of FMS were confirmed out of the 176 who were examined. The non-adjusted, adjusted, and adjusted plus weighted estimates of FMS prevalence, adjusted both for age and sex, were 2.7% (CI = 2.6%, 2.8%), 3.8% (CI = 3.7%, 3.9%), and 3.3% (CI = 3.2%, 3.4%) respectively.
Odds Ratios for Demographic variables: The likelihood of having FMS was not affected by the number of adults residing in one’s household. Less than a high school education was associated with having FMS versus chronic, non-FMS pain (compared to a university degree, OR = 3.45, CI = 2.00, 5.97) as was having an annual household income less than $12,000 (compared to an annual income of no less than $80,000, OR = 2.61, CI = 1.19, 5.71) or being disabled (compared to working or in school, OR = 2.68, CI = 1.04, 6.91).. There were increased odds of having chronic, widespread non-FMS pain (with or without FMS) among those who were divorced or separated (compared to those currently married, OR = 1.95, CI = 1.10, 3.45).
Direct Health Care Costs: Although 100 FC ultimately were identified, only 86 had been identified by closure of the Phase I telephone survey. Hence, 172 subjects were recruited to the GC group; 135 completed questionnaires were returned (78.5%). The three internal study groups (FC, PC and GC) were similar in mean age (47.8, 47.2 and 44.5 years, respectively; NS). Fewer PC were female (67% versus 86% and 82% in the FC and GC groups, respectively; p < 0.003). FC generally were less educated than GC (p < 0.05). Otherwise, the groups were demographically similar [Table 5].
Annual utilization of OHIP sponsored health services and direct health care costs billed to OHIP are presented in Table 6. Five FC, 9 PC and 41 GC refused to participate in this part of the study. The 41 GC who refused to participate were slightly younger than the 94 GC who did participate in Phase III (39.1 versus 46.0 years, p = 0.01); they were not different with respect to sex distribution, marital status, education level or household income. They also did not differ with respect to mean duration of symptoms, severity of pain or fatigue, number of symptoms, number of major symptoms, or FIQ score.
Health services and direct costs were highest for the FC group, both with respect to total annual costs and costs for each of the four cost sub-categories. Consistently, PC had the second highest annual use of services and costs, followed by paired matched subjects from the OHIP database (OC), with the GC having the lowest annual use of services and costs. The mean annual total number of health services used was 53.2 for FC, and 39.8, 28.6 and 24.9 respectively for the remaining three groups: PC, OC and GC (F = 9.46, df 255, p < 0.0001). Annual costs were $1028 (standard deviation = $1182) for FC, and $751, $536 and $463 for the other three groups, respectively (F = 8.28, df 255, p < 0.0003). In both instances, the statistical difference in group means was between the FC and GC groups. Details regarding specific services utilized also are presented in Table 6.
FC subjects reported more frequent visits to see physicians than either PC or GC subjects, over the preceding year (p < 0.0001) [Table 7]; similar results were obtained when subjects recalled the previous two weeks (p < 0.0001). They also reported more visits to physician specialists over the preceding year than the GC group (p < 0.003). Males with FMS reported more annual visits to an emergency department than males in either other group (p < 0.005). FC and PC subjects used a greater number of prescription drugs (p < 0.0001), and consumed both prescription (p < 0.0001) and over the counter (OTC) (p < 0.0001) drugs on more days over the prior two weeks than did subjects in the GC group.
The point prevalence of FMS: It has been only since 1989 that researchers have reported the prevalence of FMS in the community. Most of these early studies were in Western Europe, where the prevalence of FMS varied from 0.7% and 0.8% in Denmark24 and Finland25 to 2.0% in Germany26. Interestingly, the prevalence of FMS in Poland27 (4.5%) appears to be higher than in any Western European country. The same is true of a small town in South Africa, with a prevalence estimated at 3.2%28. A possible exception is Norway. To date, the overall prevalence of FMS in Norwegian adults has not been reported. However, FMS prevalence in young to middle-aged females is 10.5% in the small coastal town of Arendal29, and its incidence may be as high as 0.5% annually30. There are as yet no explanations for the fluctuations in FMS prevalence from country to country, and no published data as to fluctuations within countries.
Prior to 1995, estimates of the prevalence of FMS in the North American general population were based entirely upon clinic studies; these estimates ranged from 1.0%31 to 15%32. The first reported community prevalence study was performed in Wichita, Kansas33, in which the prevalence of FMS was found to be 3.2% in females, 0.5% in males, and 2.0% overall. Of 193 subjects examined, 18.9% (36) were found to meet the ACR criteria for FMS. This was only the second reported study to compare FMS cases with random community controls in terms of demographics and clinical characteristics. An earlier Danish study compared 8 confirmed cases of FMS with random controls who had 4 or fewer fibromyalgia tender points and random controls who had 5 to 10 tender points34. Both studies demonstrated clinical differences between FMS cases and controls, but lacked sufficient statistical power for many analyses.
We have been able to confirm 100 cases of FMS, in addition to 76 individuals with chronic widespread pain in whom FMS has been excluded, in a random general population survey of non-institutionalized adults in London, Ontario. We generated three estimates of FMS prevalence, both for females and males:
a conservative estimate using only confirmed cases of FMS;
an estimate including probable cases of FMS among subjects who screened positive for widespread pain but were not examined, an assumption that can be justified on the basis of it being impossible to determine whether or not FMS itself enhanced or diminished an eligible subject's likelihood to participate in Phase II, and our own data that shows that Phase II non-participants were similar to participants, both demographically and symptomatically;
an estimate to adjust for over-representation by seniors in the sample.
Despite using quite different assumptions, the three estimates were very similar. An intermediate estimate of FMS prevalence among non-institutionalized females is 4.9% (4.8%, 4.9%) and 1.6% in males (1.5%, 1.8%), for a female to male prevalence ratio of approximately three to one. Among non-institutionalized adults in the general community, the percentage FMS cases that are female is 75.0% (72.5%, 76.9%). This percentage is at the low end of what generally has been reported in clinic studies of FMS35. Roughly one in twenty adult females and one in sixty adult males currently have FMS. Besides female sex, risk factors for having FMS versus chronic pain from another source include middle age, low household income, and less than a high school education. These four risk factors also were identified in Wichita36.
Assuming that London is representative of the adult population of Canada, then FMS currently affects almost 700,000 Canadian adults. FMS is approximately four to eight times as common as rheumatoid arthritis37, and much more common than systemic lupus38.
FMS generally is regarded to be a non-infectious, chronic, non-remitting, non-crippling and non-fatal disorder39. This characterization of FMS is not supported by our data, especially in females, for whom FMS prevalence steadily rises from age 18 through middle age, then appears to steadily decline. Weighting according to the number of adults residing in each household results in a reduction in the slope from early to late middle age, probably by reducing the bias towards over-counting elderly subjects living alone. Nonetheless, the peak FMS prevalence in the 55 to 64 year old age group is maintained.
There are at least three potential explanations for the observed age effect. The first is that FMS does remit, particularly in individuals over age 64. In the longest, prospective study of FMS clinic patients, a majority reported some improvement over ten years, but no cases of complete remission were noted40. In an eight year Swedish study of 49 FMS patients, remissions were described as ‘rare’41.
A second explanation is selective mortality; in other words, individuals with FMS are more likely to die than age and sex matched individuals in the general population. FMS either could have a potentially fatal course, or it could be associated with other, potentially fatal illnesses. Clinic studies already have demonstrated that FMS is commonly seen in patients with systemic lupus42 and rheumatoid arthritis43, diseases for which the standardized mortality ratio (SMR) is greater than one.
A third hypothesis is that there is a cohort effect (an effect similar to that of the ‘baby boom’ generation). Sometime in the past, there may have been an epidemic of FMS, specific to one particular age group. If FMS truly is chronic, non-remitting and non-fatal, then this peak of FMS prevalence would follow this cohort as it ages. Ten years from now, we would expect a fall in FMS prevalence in the 55 to 64 year age group, and a rise in prevalence in individuals 65 to 74. Allowing for this possibility, then the ‘epidemic of FMS’ must be explained, and the potential especially for an infectious or other precipitating cause must be re-examined.
It is possible that our study was biased towards confirming FMS in middle-aged, rather than older individuals, if older persons with FMS are less willing or able to participate either in the initial screening survey or the confirmatory physical examination. However, several findings make this unlikely. First, the representation of people over age 65 in the sample (15.2%) is very close to the proportion of the population in this age group (15.5%). Second, while the proportion of positive screens who participated in Phase II is lower among those 65 and older (58%) than among those under 65 years (75%), this lower participation rate is insufficient, by itself, to account for the difference in prevalence observed after age 64.
One might argue that the fall in FMS prevalence commencing at age 65 supports the claim that FMS is a product of an over-generous compensation system. This explanation is not supported by the higher prevalence of FMS in Poland and South Africa than in Western Europe and the U.S. Nor is it supported by the gradual decline in FMS prevalence, approximately three percent per decade after age 64, rather than a sharp decline at age 65.
It is possible that the apparent decline of FMS prevalence in those over age 64 is a result of a bias towards selectively surveying well elders at home, as opposed to more ill elders in institutions. In 1991, 6300 Londoners,1.8% of the population, did not reside in private households. We can estimate the effect of institutionalized adults on FMS prevalence on females age 65 and older. Assume that half of the 6300 institutionalized persons in London are female, and that 50% of these females are 65 or older. If 10% of them have FMS (double the rate in the non-institutionalized) it would inflate the crude estimate of FMS prevalence in women over 64 from 3.6% (95% CI = 2.7%, 4.8%) to 3.7% (95% CI = 2.8%, 4.9%). This compares to a crude prevalence of 8.5% in the 55 to 64 age group (95% CI = 6.8%, 10.6%). Hence, even correcting for a prevalence of FMS among institutionalized females that is double that of non-institutionalized females, there is a statistically significant decline in prevalence in women 65 and older.
Self-reported medication and health service utilization: FMS cases reported using more medications and health services than any of three control groups, including those with chronic widespread pain without FMS. The differences in medication use was both for prescribed and over-the-counter medication. The difference in self-reported health services use primarily centered around physician services, rather than services by other health professionals. These differences were present both for one year and two week recall. Most of the services utilized were provided on an outpatient basis, the average individual with FMS spending only two days as a hospital inpatient per year.
It is somewhat surprising that males, not females with FMS, utilized emergency department services more than did controls. It may be that males were less likely to utilize the services of their family physicians; males with FMS reported somewhat fewer visits to their family physicians over the preceding year (6.9 versus 9.6 visits for females). More numerous or more severe symptoms might lead certain patients to seek emergency care, rather than the care of their family physician, but there were no sex differences in symptom quantity or severity.
Direct health care reimbursement costs: Direct health care costs incurred by the Ontario Health Insurance Plan (OHIP) were higher for FMS cases than any of the three control groups, including those with chronic widespread pain without FMS. The difference between annual reimbursement costs for FMS cases versus general population controls from the OHIP database was $493 per annum. Multiplied by 687,000 adults with FMS Canada wide, this represents almost $350 million in direct health care costs across Canada in 1993. The lion’s share of this was for physician services, approximately $250 million annually. The estimated reimbursements for other health professionals, laboratory and radiology services were $21 million, $41 million, and $23 million , respectively. These costs are only a percentage of the actual costs of FMS, which would include other direct costs such as those for clinic staffing and maintenance, and indirect costs such as lost income and insurance pensions, the last of which has been estimated as high as $200 million annually44. Data from the U.S. 1990-1992 National Health Interview Survey (NHIS) suggest that somewhat less than 50% of the costs of musculoskeletal illness are direct health care costs45. If this is true for FMS, then the annual cost of FMS in Canada likely exceeds $700 million.
In summary, we found FMS to be a widely prevalent illness which results in substantial expenses, at least to the health care system. Irrespective of whether one views it as a legitimate medical condition or a medicalization of a social phenomenon, the frequency and costs of FMS suggest the need for further investigations into its etiology and treatment.
This research was supported by grants from the Arthritis Society of Canada, the National Health Research and Development Program of Canada, and the London Health Sciences Centre. Thanks also to Mr. Larry Stitts, Dr. James Rochon, and Dr. John Koval for advice on statistical analyses, to Ms. Brenda Nesbitt for her assistance in the preparation of this manuscript, and to all of the telephone interviewers who participated in this project.
List of Figures and Tables
The LFES screening questionnaire (LFES-SQ).
Confirmed and estimated female cases of FMS in the survey sample.
The prevalence of FMS in non-institutionalized adult females - London, Ontario
Confirmed and estimated male cases of FMS in the survey sample.
The prevalence of FMS in non-institutionalized adult males - London, Ontario.
Demographic comparison of FMS cases (FC), Pain controls (PC) and General controls (GC).
Services reimbursed and amount paid in 1993 by the Ontario Health Insurance Plan (OHIP).
One year utilization of health services by group: FC, PC, HC.
Figure 1: The LFES Screening Questionnaire (LFES-SQ)
In the past three months, have you had pain in muscles bones or joints, lasting at least one week?
Have you had pain in your shoulders, arms or hands? On which side? Right, left or both?
Have you had pain in your legs or feet? On which side? Right, left or both?
Have you had pain in your neck, chest or back?
Meeting the pain criteria requires `yes' responses to all four pain items, and either
both a right and left side positive response, or 2) a both sides positive response.
1 Lawrence RC, Hochberg MC, Kelsey JL, et al. Estimates of prevalence of selected arthritic and musculoskeletal diseases in the United States. J. Rheumatol, 1989;16:427-441.
2 Felts W, Yelin E: The economic impact of the rheumatic diseases in the United States. J Rheumatol, 1989; 16: 867-884.
3 Wood PHN, Badley EM: Rheumatic Disorders. In: Miller DL, Farmer RDT, eds. Epidemiology of Diseases. Oxford: Blackwell Scientific Publications, 1982: 333-346.
4 Reynolds DL, Chambers LW, Badley EM, Bennett KJ, Goldsmith CH, Jamieson E, Torrance GW
, Tugwell P: Physical Disability among Canadians Reporting Musculoskeletal Diseases. J Rheumatol 1992;19 (7):1020-1030.
5 Lee P, Helewa A, Smythe HA, Bombardier C, Goldsmith CH: Epidemiology of Musculoskeletal Disorders (Complaints) and Related Disability in Canada. J Rheumatol, 1985;12:1169-1173.
6 Badley EH, Yoshida K, Webster G, Stephens M: Ontario Health Survey 1990, Working Paper No. 5: Disablement and Chronic Health Problems in Ontario. Ministry of Health Ontario, 1993.
7 Spitzer WO, Harth MH, Goldsmith CH, Norman GR, Dickie GL, Bass MJ, Newell JP: The Arthritis Complaint in Primary Care: Prevalence, Related Disability, and Costs. J Rheumatol,1976;3 (1):88-99.
8 Felts W, Yelin E: The Economic Impact of the Rheumatic Disorders in the United States. J Rheumatol, 1989;16:867-884.
9 Yelin E, Callahan LF: The economic cost and social and psychological impact of musculoskeletal conditions. Arthritis Rheum, 1995; 38 (10): 1351-1362.
10 Wolfe F: The Clinical Syndrome of Fibrositis. Amer J Med, 1986;81 (3A):7-14.
11 Hawley DJ, Wolfe F: Pain, disability, and pain/disability relationships in seven rheumatic disorders: A study of 1,522 patients. J Rheumatol, 1991; 18 (10): 1552-57
, Martinez JE
, Ferraz MB, Sato EI, Atra E: Fibromyalgia versus rheumatoid arthritis: A longitudinal comparison of the quality of life. J Rheumatol, 1995; 22 (2): 270-74
, Burckhardt CS, Clark SR, Bennett RM: Fibromyalgia and quality of life: A comparative analysis. J Rheumatol, 1993; 20 (3): 475-79.
. Bruusgaard D, Evensen AR, Bjerkedal T: Fibromyalgia - A New Cause For Disability Pension. Scand J Soc Med, 1993;21:116-119.
, Ochoa JL: Essence, Investigation, and Management of "Neuropathic" Pains: Hopes from Acknowledgement of Chaos. Muscle Nerve, 1993;16:977-1008.
, Hadler NM: Occupational Musculoskeletal Disorders. Raven Press, New York, 1993:16-33.
12 Wallace DJ: Systemic Lupus Erythematosis, Rheumatology and Medical Literature: Current Trends. J Rheum, 1985; 12:913-15.
13 Marder WD, Meenan RF, Felson DT, Reichlin M, Birnbaum NS, Croft JD, Dore RK, Kaplan H, Kaufman RL, Stobo JD. The Present and Future Adequacy of Rheumatology Manpower. Arthritis Rheum, 1991;34:1209-1217.
14 Alarcon-Segovia D
, Ramos-Niembro F, Gonzales-Amaro RF. One Thousand Private Rheumatology Patients in Mexico City. Arthritis Rheum, 1983;26:688-689 [letter].
15 Calabozo-Raluy M, Llamazares-Gonzales AI, Munoz-Gallo MT, Alonso-Ruiz A: Sindrome de Fibromialgia (Fibrositis); Tan Frequente Como Desconcido. Med Clin Barc, 1990;94:173-175.
16 Reilly PA, Littlejohn GO: Peripheral Arthralgic Presentation of Fibrositis/Fibromyalgia Syndrome Rheumatol, 1992;19:281-283.
17 White KP, Harth M, Speechley M, Ostbye T: Fibromyalgia in Rheumatology Practice: A Survey of Canadian Rheumatologists. J Rheumatol, 1995;22:722-726.
18 Wolfe F, Ross K, Anderson J
, Russell IJ, Hebert L. The Prevalence and characteristics of fibromyalgia in the general population. Arthritis Rheum, 1995; 38:19-28.
19 O’Rourke D, Blair J: Improving random respondent selection in telephone surveys. Journal of Marketing Research 1983; 20: 428-432.
20 Badley EH, Yoshida K, Webster G, Stephens M: Ontario Health Survey, 1990. Working Paper Number 5. Disablement and chronic health Problems in Ontario. Health Canada.
21 Wolfe F, Smythe HA, Unus MB, et al: The American College of Rheumatology 1990 criteria for the classification of fibromyalgia. Report of the multicenter criteria committee. Arthritis Rheum 1990;33:160-172.
22 Scheaffer RL, Mendenhall W, Ott L: Elementary Survey Sampling, Fourth Edition. PWS-Kent Publishing Co., Boston, MA, 1990.
th Rochon J: The application of the GSK method to the determination of minimum sample sizes. Biometrics, 1989; 45: 193-205.
23 SAS Release 6.12, 1996. SAS Institute Inc., Cary, North Carolina
24 Prescott E, Kjoller M, Jacobsen S, Bulow PM, Danneskoild-Samsoe B, Kamper-Jorgensen. Fibromyalgia in the adult Danish population. Scand J Rheumatol, 1993; 22: 233-37.
25 Makela M, Heliovaara M: Prevalence of primary fibromyalgia in the Finnish population. BMJ, 1991; 303 (7): 216-19.
26 Raspe H, Baumgartner Ch. The epidemiology of the fibromyalgia syndrome (FMS) in a German town. Scand J Rheumatol, 1992; suppl 94: 8.
27 Schochat T, Croft P, Raspe H: The epidemiology of fibromyalgia. Workshop of the standing committee on epidemiology, european league against rheumatism (EULAR), Bad Sackingen, 19-21 November 1992. Brit J Rheumatol, 1994; 33: 783-786.
28 Lyddell C, Meyers OL: The prevalence of fibromyalgia in a South African community. Scand J Rheumatol, 1992; suppl 94: 8.
29 Forseth KO, Gran JT: The prevalence of fibromyalgia among women aged 20-49 years in Arendal, Norway. Scand J Rheumatol, 1992; 21: 74-78.
30 Forseth KO, Gran JT, Husby G: A population study of the incidence of fibromyalgia among females aged 26 - 55 years. Arthritis Rheum, 1997; 40 (9) [suppl]: S44.
31 Wallace DJ: Systemic Lupus Erythematosis, Rheumatology and Medical Literature: Current Trends. J Rheum, 1985; 12:913-15.
32 Gerecz-Simon EM, Kean WF, Watson-Buchanon W: New criteria for the classification of fibromyalgia. Can J Diagnos, February 1990: 80-96.
33 Wolfe F, Ross K, Anderson J, Russell IJ, Hebert L: The prevalence and characteristics of fibromyalgia in the general population. Arthritis Rheum, 1995; 38 (1): 19-28.
34 Prescott E, Jacobsen S, Kjoller M, Bulow PM, Danneskold-Samsoe B, Kamper-Jorgensen F: Fibromyalgia in the adult Danish population: II. A study of clinical features. Scand J Rheumatol, 1993; 22 (5): 238-42.
35 Goldenberg D: Fibromyalgia syndrome. An emerging but controversial condition. JAMA, 1987; 257 (20):
36 Wolfe F, Ross K, Anderson J, Russell IJ, Hebert L: The prevalence and characteristics of fibromyalgia in the general population. Arthritis Rheum, 1995; 38 (1): 19-28.
37 Spector TD: Epidemiology of Rheumatic Diseases: Rheumatoid arthritis. Rheum Dis Clin N Amer, 1990; 16 (3): 513-537.
38 Schur PH: Clinical features of SLE, in Textbook of Rheumatology, 3rd Ed., EditorsKelley WN, Harris ED Jr, Ruddy S, & Sledge CB, 1989. WB Saunders Co., Harcourt Brace Jovanovich, Inc., Philadelphia, USA.
39 Wolfe F and the Vancouver fibromyalgia consensus group. The fibromyalgia syndrome: A consensus report on fibromyalgia and disability. J Rheumatol, 1996; 23 (3): 534-539.
40 Kennedy M, Felson DT: A prospective long-term study of fibromyalgia syndrome. Arthritis Rheum, 1996; 39 (4): 682-685.
41 Bengtsson A
, Backman E: Long-term follow up of fibromyalgia patients. Scand J Rheumatol, 1992; suppl 94:9 8.
42 Romano TJ: Coexistence of fibromyalgia syndrome (FMS) and systemic lupus erythematosis. Scand J Rheumatol, 1992; suppl 94: 12.
43 Romano TJ: Incidence of fibromyalgia syndrome (FMS) in rheumatoid arthritis (RA) patients seen in a general rheumatology practice. Scand J Rheumatol, 1992; suppl 94: 11.
44 McCain G, Cameron R, Kennedy JC: The problem of longterm disability payments and litigation in primary fibromyalgia: The Canadian perspective. J Rheumatol, 1989; 16 (suppl 19): 174-76.
45 Yelin E, Callahan LF: The economic cost and social and physical impact of musculoskeletal conditions. Arthritis Rheum, 1995; 38 (10): 1351-62.