Complementary and alternative medicine (CAM) is

original research UTILIZATION, PATIENT SATISFACTION, AND COST IMPLICATIONS OF ACUPUNCTURE, MASSAGE, AND NATUROPATHIC MEDICINE OFFERED AS COVERED HEALTH BENEFITS: A COMPARISON OF TWO DELIVERY MODELS Darrell Stewart, MHA, John Weeks, and Stephen Bent, MD Darrell Stewart is president of Advantis in Tucson, Ariz, and was president of Alternáre Health Services in Seattle, Wash, when this article was written. John Weeks is the editor of THE INTEGRATOR for the Business of Alternative Medicine in Seattle, Wash. Stephen Bent is assistant professor of medicine with the Osher Center for Integrative Medicine at the University of California, San Francisco. Context Increasing numbers of health plans in the United States offer complementary and alternative medicine (CAM) benefits despite limited information. Objective To determine the utilization rates and costs associated with providing CAM services in 2 benefit designs, and to determine the satisfaction of patients in both plans. Design Two health plans were identified: a traditional indemnity plan offered through a defined preferred provider organization (PPO) of CAM providers and a health maintenance organization (HMO). Costs and utilization rates for CAM services were compared during a 1-year period of coverage beginning November 1, 1996. Setting and Participants 1091 patients in both plans who used CAM services during the month of May 1997 in Washington state. Results Only 1% of all patients covered for CAM accessed these services during the study period. A significantly higher percentage of patients in the PPO plan (1.2%) used CAM services compared to the HMO plan (0.6%) (P<.001). However, the average total cost of annual CAM services (plan benefit + user contribution) was similar ($347 in the HMO and $376 in the PPO), and the price per member per month was nearly identical ($0.20 in the HMO and $0.19 in the PPO). Most users perceived these services as helpful. Conclusions Utilization of CAM services and per member per month costs were lower than expected given the high interest in CAM services reported in consumer surveys. The high level of satisfaction with CAM services and self-reported decrease in the use of pain medications suggests the need for prospective studies examining the effect of CAM treatments. (Altern Ther Health Med. 2001;7(4):66-70) Reprint requests: InnoVision Communications, 169 Saxony Rd, Suite 104, Encinitas, CA 92024; phone, (760) 633-3910 or (866) 828-2962; fax, (760) 633-3918; e-mail, alternative.therapies@ innerdoorway.com. Complementary and alternative medicine (CAM) is becoming increasingly popular among consumers of healthcare. 1-3 Some health plans in the United States have responded to this consumer demand and have begun to offer a variety of CAM benefits. 4-6 Most health plans already offer the option of chiropractic care, 7 but many are now experimenting with newer benefit packages that include services of acupuncturists, massage practitioners, and naturopathic physicians in states where these providers are licensed. 8-11 Surveys of consumers have shown significant interest in these services where they are available. 1-3,9 Health plans are beginning to offer these benefits with limited information about expected utilization patterns or associated costs. 12 A recent Society of Actuaries study of cost and utilization in alternative care did not find any significant data on which an actuary could base premium rates (oral communication, Lee Launer, March 1998). Estimates of actuaries contracted by a health plan to assist in pricing services of these providers have been reported to be off by as much as a factor of 10. 13,14 Health plans therefore face difficulties in determining the best way to structure, deliver, and price these CAM benefits. We are unaware of any published data on per member per month (PMPM) costs for offering services of acupuncturists, massage practitioners, and naturopaths through credentialed networks of providers. Whereas consumer satisfaction with CAM services has been shown, 3 satisfaction data on users of these services when provided through covered benefits are not available. In addition, we are aware of no literature that examines patients perceptions of the effect of CAM services on their subsequent use of pain medication and conventional medical care. Washington state provides a unique opportunity to study CAM services provided through credentialed networks. In 1995, the Washington state legislature passed a mandate that required all plans to include every (licensed) category of provider as of January 1, 1996. 15 The health plans in the state with the largest enrollment all met the mandate by offering benefits through credentialed provider networks for acupuncture, massage therapy, and 66 ALTERNATIVE THERAPIES, july/aug 2001, VOL. 7, NO. 4 Delivery Models for Acupuncture, Massage, Naturopathy

naturopathic medical treatment. The mandate and methods of compliance have effectively made Washington state a laboratory for the study of the inclusion of CAM into covered benefit packages. In this study, we first compare the utilization patterns for CAM under 2 benefit designs: a traditional indemnity benefit (PPO) and a health maintenance organization benefit (HMO), both offered exclusively through the same network of contracted CAM providers. We also report the results of a survey of all patients who received CAM services from both benefit plans during a 1-month study period. The survey includes patient perceptions on whether these CAM services impacted their use of conventional services and prescription drugs. Our investigation provides an initial examination of the utilization rates, costs of providing services, and satisfaction rates for patients receiving acupuncture, naturopathic medical treatment, and massage therapy when these services are offered as covered health benefits. METHODS Comparison of Benefit Plans We first identified 2 disparate CAM benefit designs operating in Washington state during 1997. One was a traditional indemnity plan (Premera-Blue Cross, Mountlake Terrace, Wash) offering CAM services through a credentialed network (PPO) that included the following features: (1) direct access by plan members to acupuncturists, naturopathic physicians, and massage practitioners; and (2) 50% of allowable charges paid by the consumer as coinsurance on a maximum annual benefit of $1000 ($500 paid by insurer) for combined acupuncture, massage, naturopathic, chiropractic, and nutritional counseling services. A small percent of individuals examined in this study, represented by roughly 10 small groups of 50 to 100 lives (up to 0.2% of the covered individuals), had a benefit with a smaller copay (usually 20%) and higher maximum (usually $1000). The second plan was managed, offered by an HMO (Group Health Cooperative of Puget Sound, Seattle, Wash), and included the following: (1) referral from a conventional primary care provider was required; (2) the maximum number of visits per referral was 3 for naturopathic medical treatment, 6 for massage therapy, and 6 for acupuncture, with additional visits subject to review; (3) coverage was limited to a specified set of conditions for each provider type; (4) reporting to the HMO and the conventional provider by the CAM provider was required; (5) no predetermined limit on the total allowable services; and (6) no co-payment required by the patient. Conditions selected for coverage by the HMO were developed out of a process that included a review of the literature and information provided by representatives of the 3 professions. 16 Both plans contracted with the identical credentialed network (Alternáre of Washington, a subsidiary of The Alternáre Group, Inc, based in Portland, Ore) for the provision of acupuncture, naturopathy, and massage therapy. Payments to providers were based upon the same established fee schedule. Services provided outside of the contracted, credentialed network were not covered by either plan. Financial risk related to CAM service utilization was borne by the plans and not extended to primary care physicians. We obtained data from each plan s claims payment system on the total number of visits to CAM providers, total costs of CAM services provided, proportion of eligible beneficiaries using CAM services, average costs for CAM services, and monthly trend data for the 1-year period from November 1, 1996, to October 31, 1997. The data sources were subject to routine audits for accuracy. The proportion of eligible beneficiaries using services ( users ) between plans was compared using the chi-square test. At the time of this study, neither plan offered primary care physicians reimbursement for CAM services. Both plans offered chiropractic benefits through existing contracted networks, but chiropractic care was not included in this analysis. Although some naturopathic physicians also practiced chiropractic medicine, only their naturopathic services were considered in this analysis. User Survey The consumer survey consisted of a set of questions that was mailed to all users of covered services of naturopaths, acupuncturists, and massage practitioners under these 2 plans during the month of May 1997. The mailing list was determined from records of the alternative network management firm contracted by the 2 plans. Questions included perceived helpfulness of the CAM services offered by these providers, perceived impact on use of the conventional team, and perceived impact on the use of prescription drugs. The instrument included an introductory letter that noted the name of the health plan and the name of the user s CAM provider(s). Patient names were not recorded on the form, but a unique numeric code allowed for analysis of demographic data. Codes were kept confidential. Surveys were mailed on August 22, 1997, to the 1091 identified users of CAM services who had been seen in May 1997 and for whom a claim had been received by the network management firm as of August 15, 1997. A second mailing to nonrespondents was sent on October 28, 1997. A total of 89 surveys (8%) were returned as undeliverable, leaving 960 possible respondents. Response rates to different questions were compared using the chi-square test. RESULTS Utilization Data The data on overall utilization for the 1-year period from November 1, 1996, to October 31, 1997, are reported in Table 1. Utilization of services under both plans was considerably less than might be expected from consumer surveys of member interests by health plans. Overall, only 1% of the covered individuals in the 2 plans accessed covered CAM services from the 3 provider categories under investigation during the study period. A significantly higher percentage of patients in the PPO plan (1.2%) accessed CAM services than did those in the HMO plan (0.6%) (P<.001). The cost to the health plans for covering these services, relative to the low utilization and total claims paid for all medical services, was correspondingly low (Table 1). The gross charges reflected just $2.30 per member per year on average, or $0.19 PMPM. Although a higher percentage of PPO insureds used the services, the PMPM expense in the HMO benefit was marginally Delivery Models for Acupuncture, Massage, Naturopathy ALTERNATIVE THERAPIES, july/aug 2001, VOL. 7, NO. 4 67

TABLE 1 CAM utilization experience in an HMO and PPO plan* Measured HMO: referral only, no cap on services PPO: direct access/ capped benefit payments, $ Average enrollment in plan during study Cost per member per year, $ Cost per member per month, $ members/insured using CAM benefits % of eligible using CAM benefits Average payment per user, $ % of 12-month activity in first 3-month period % of 12-month activity in last 3-month period 883305 374833 2.36 0.20 2540 0.6 347 23 27 1043991 461392 2.26 0.19 5550 1.2 188 21 27 1927296 836225 2.30 0.19 8090 1.0 238 NA NA * CAM indicates complementary and alternative medicine; HMO, health maintenance organization; and PPO, preferred provider organization. From November 1, 1996, through October 31, 1997. P<.001 for comparison of percent of eligible using CAM benefits between plans. P<.001 for comparison first 3 months to last 3 months (within both plans). higher ($0.20 in the HMO and $0.19 in the PPO). This reflected higher average cost of claims paid per individual user in the HMO ($347) during the 12-month period compared to cost to the insurer for users of the PPO plan benefits ($188). The average total cost of annual CAM services (plan benefit plus user contribution) for these 3 provider types, among users, was $300 to $400 in both plans ($347 plus minimal co-payments in the HMO; $376 in the PPO, if the plan s exposure [$188] is doubled to reflect coinsurance). An estimate of trends in costs was measured by comparing data on the first 3-month period and final 3- month period of the 12 months analyzed. Utilization rose in each plan and had not stabilized. User Survey A total of 675 patients responded to the survey, for an overall response rate of 70%. Response rates for the 2 sets of users of acupuncture, massage, and naturopathic medical services were similar (72% HMO, 69% PPO). Demographic data for the users of CAM services in May 1997 are offered in Table 2. Users under both benefit designs were aged 40 to 60 years. A significantly higher percentage of patients with direct access to covered CAM services under the traditional indemnity plan were aged less than 41 years (43%) compared to users receiving referrals under the HMO benefit (26%) (P<.001). Both plans showed a higher percentage of female patients, with overall percentages at 78% female and 22% male. Overall findings from the survey are presented in Table 3. Most users of covered acupuncture, massage, and naturopathic services perceived these services as extremely (56%) or very (26%) helpful. Another 10% marked moderately helpful. For the same group, a total of 44% found the services of their traditional care to be at least moderately helpful. A much higher percentage of patients was at least moderately satisfied with their care from these CAM providers (92%) compared to their conventional care (44%) (P <.001). For those ranking the helpfulness above moderate, the figures were 82% for CAM care and 17% for conventional care. When asked whether they would do it over again, more than three fourths (76%) answering the survey indicated that they definitely would, and another 13% indicated they probably would see the CAM provider again. Three consumers (0.4%) reported that the CAM services offered through these credentialed networks of acupuncturists, massage practitioners, and naturopathic physicians had been harmful. More than half of the respondents believed that the experience decreased their use of conventional care slightly or substantially. This finding was consistent whether the question related to the use of conventional team (55%) or of prescription drugs (61%). Very few respondents perceived that the impact of these CAM services increased their use of conventional services (2%) or prescription drugs (1%). DISCUSSION To our knowledge, this is the first study of resource utilization and patient satisfaction for CAM services provided by health plans through credentialed networks of acupuncturists, massage practitioners, and naturopathic physicians. The picture that emerges is of unexpectedly low utilization, trending upward over time, coupled with high patient satisfaction with CAM services. We anticipate that these findings will be viewed with interest by health plans investigating the coverage of CAM therapies. The low utilization of CAM services was surprising given the high interest in CAM found in surveys. 1-3 One survey by a health plan in the similar CAM market of Portland, Ore, found that 21% of members were very interested and 26% somewhat interested (47% in both categories) in using services of 68 ALTERNATIVE THERAPIES, july/aug 2001, VOL. 7, NO. 4 Delivery Models for Acupuncture, Massage, Naturopathy

naturopathic doctors; corresponding findings for acupuncturists were 14% and 25% (39% in both). 9 Whereas a national telephone consumer survey in 1997 found very low utilization of acupuncture (2%) and naturopathy (1%), 3 it is reasonable to expect a much higher utilization rate in Washington given the higher availability of these services in this state relative to the rest of the country. The same national survey concluded that massage therapists, a nationally available provider category, were used by 14% of those interviewed. The utilization of massage therapists on a national level is therefore 14 times the utilization of all CAM services (acupuncture, massage therapy, naturopathy) by members of the health plans included in this study (which showed an overall utilization of 1%). Why did we observe such a low utilization of these services when surveys have shown such high public interest in CAM? Answering this question was beyond the scope of our study, but an informal report on the experience of all of Washington s major plans that followed the first year of the mandate advanced several possible reasons for this finding. 13,14 These included (1) medical necessity requirements that denied coverage when these providers were used for other purposes, such as general wellness; (2) low knowledge of these CAM services among primary care physicians, creating few referrals that were not requested by patients; (3) the low profile given the CAM benefits by most plans, thereby limiting awareness of the benefit among policy holders; (4) benefit design restrictions on the number of CAM providers who could be accessed; (5) limits on the conditions for which services would be covered; and (6) perception among some consumers that gaining coverage might be more difficult than it was worth. Notably, all but one of the medical directors interviewed for the report expected that utilization of CAM services would increase over time. 13,14 The expectation of increased utilization and cost was supported by our study. Age, y 0-20 21-40 41-60 61-80 > 81 Female Male TABLE 2 Demographics of users of CAM services* HMO 7 2 105 24 269 61 51 12 6 1 438 100 349 80 89 20 PPO 30 6 194 37 258 49 39 7 2 522 100 391 75 131 25 41 4 355 33 591 54 96 9 8 1 1091 100 848 78 243 22 * Data for all users of covered benefits delivered by acupuncturists, massage practitioners, and naturopathic physicians in 2 benefit designs during May 1997. CAM indicates complementary and alternative medicine; HMO, health maintenance organization; and PPO, preferred provider organization. This study suggests that health plans experimenting in offering CAM services through credentialed networks of acupuncturists, massage practitioners, and naturopathic physicians may have substantial flexibility in their efforts to control costs while implementing managed CAM benefits. We note the finding of a nearly identical PMPM price for these services under these disparate benefit designs. In addition, the total annual cost of CAM services per user was similar under both benefit designs: between $350 and $400 per user. Users of CAM benefits in the open-access PPO model did not come close to spending their maximum benefit ($188 average annual cost to the insurer per CAM user out of a maximum $500 benefit). Even though CAM users are reporting high rates of satisfaction, they are not spending up to the limit of their benefit plan. PERCEPTION OF HELPFULNESS OF CAM SERVICES Our study found that services of these CAM providers met with high satisfaction among users. The satisfaction rate compares favorably with results of surveys of users who may be paying wholly out of pocket. A commercially sponsored survey in 1997 reported that 80% of users of massage services and 57% of users of acupuncture services found them very helpful. 3 Our findings were 85% and 81%, respectively, for users of these 2 categories. Most users of CAM treatments in this study reported that their use of conventional services and prescription drugs decreased after their use of CAM. Whereas this finding does not prove that Measured TABLE 3 Responses of users of covered CAM benefits* Satisfaction with services Care from conventional physician for the condition was very (17%) to moderately (27%) helpful Care from CAM provider was extremely (56%), very (26%) or moderately (10%) helpful CAM care was harmful Would definitely (76%) or probably (13%) return to the same CAM provider with the same condition if they had to do it all over again Impact on conventional care Use of conventional team perceived as having decreased slightly (30%) or substantially (25%) Use of prescription drugs perceived as having decreased slightly (27%) or substantially (34%) Increased use of conventional team Increased use of prescription drugs Result, % 44 92 0.4 89 * 1091 surveys mailed with 89 returned to sender, yielding 960 surveys; 675 respondents (70%). CAM indicates complementary and alternative medicine. 3 individuals (0.4%). 55 61 2 1 Delivery Models for Acupuncture, Massage, Naturopathy ALTERNATIVE THERAPIES, july/aug 2001, VOL. 7, NO. 4 69

CAM treatment reduces the cost of conventional care, it suggests that there may be cost offsets with the potential for lower overall costs of care for certain conditions. Although patient perceptions of reduced use of prescription drugs and conventional services is interesting, this finding should be confirmed with studies that include an examination of pharmacy records, conventional service utilization, and claims data. If reduced use of conventional services is confirmed for users of CAM, the provision of CAM services may be cost-neutral or even cost-saving for insurers. Ideally, randomized controlled trials of specific CAM therapies for precise indications are needed to definitively examine the effect of CAM treatment on all patients, not merely those who have either personally chosen to use CAM or who were referred by their provider. Our study has several limitations. The information on utilization was obtained in Washington state after a legislative mandate was issued requiring coverage for acupuncture, naturopathy, and massage therapy. It is not clear that the utilization and cost information will be applicable to other states in which coverage is not mandated, but independently developed, advertised, and provided by an HMO or an insurance plan. Although the observed utilization rates in our study were low, we found an increase in utilization during the study period and cannot predict how this utilization will change over time. The survey of patient satisfaction was offered only to patients who received CAM benefits. Although these patients were very satisfied with their care, we do not know the satisfaction rates of patients who received only conventional care for similar medical conditions. Therefore, the survey suggests that CAM benefits are well received by patients who choose to use them. In this group of patients, the use of CAM may result in lower use of conventional services and prescription drugs. The early experience of these 2 large health plans in Washington state suggests that the utilization of CAM benefits and costs associated with these benefits are low. Two entirely different models of delivering CAM benefits resulted in similar costs PMPM. Users of CAM are clearly very satisfied with these treatments, and more than half report that the use of CAM results in a reduction in the use of conventional care and prescription drugs. The effect of CAM on disease-specific outcomes and the overall cost of care requires further study. This will best be accomplished through the careful development of randomized controlled trials of CAM treatments for specific medical conditions, and will clearly require substantial time and funding. In the meantime, health plans providing CAM benefits can be reassured that current users of CAM are very satisfied with this type of care when it is provided as a covered benefit. Whereas overall costs of these services are not yet known, patients report that the use of CAM leads to a reduction in their use of conventional medical care. References 1. Eisenberg DM, Kessler RC, Foster C, Norlock FE, Calkins DR, Delbanco TL. Unconventional medicine in the United States: prevalence, costs, and patterns of use. N Engl J Med. 1993;328:246-252. 2. Eisenberg DM, Davis RB, Ettner SL, et al. Trends in alternative medicine use in the United States, 1990-1997. JAMA. 1998;280:1569-1575. 3. Landmark Healthcare. The Landmark Report on Public Perceptions of Alternative Care. Sacramento, Calif: Landmark Healthcare; 1998. 4. Pelletier KR, Marie A, Krasner MA, Haskell WL. Current trends in the integration and reimbursement of complementary and alternative medicine by managed care, insurance carriers, and hospital providers. Am J Health Promot. 1997;12:112-123. 5. Stern D. The building blocks of a complementary medicine program. Healthplan. January/February 1998:63-68. 6. Weeks J. 1998 CAM network executive survey, part I. THE INTEGRATOR for the Business of Alternative Medicine. Seattle, Wash: Integration Strategies for Natural Healthcare. December 1998:1,3-7. 7. Gray P. Chiropractic: the new benefit staple. Business and Health. March 1998:45. 8. Goch L. Alternative medicine moves into the mainstream. Best s Review L/H. March 1997:86. 9. Weeks J. Oregon survey on CAM interest. THE INTEGRATOR for the Business of Alternative Medicine. Seattle, Wash: Integration Strategies for Natural Healthcare. May 1997:7. 10. Hoover E. Alternative care gains fiscal footing. The Oregonian. April 29, 1998:B1. 11. Lippman H. Executive opinion poll. Business and Health. December 1997:36-41. 12. Weeks J. Estimating acupuncture utilization and costs: a (necessarily) impressionistic portrait. THE INTEGRATOR for the Business of Alternative Medicine. Seattle, Wash: Integration Strategies for Natural Healthcare. December 1998:5. 13. Weeks J. Marrying the medicines. Seattle Times-Post Intelligencer. April 6, 1997:F1. 14. Weeks J. Special report on Washington s every category mandate, part one. THE INTE- GRATOR for the Business of Alternative Medicine. Seattle, Wash: Integration Strategies for Natural Healthcare. June 1997:4. 15. Revised Code of Washington 48.43.045. 16. Weeks J. Structure for CAM integration. In: Kongstvedt PR, Ploscher DW, eds. Best Practices in Medical Management. Gaithersburg, Md: Aspen; 1998:605-608. Acknowledgments The authors acknowledge the valued assistance of Laura Patton, MD, medical director of Alternative Services, Group Health Cooperative of Puget Sound; and Peter West, MD, associate medical director of Premera/Blue Cross of Washington and Alaska. 70 ALTERNATIVE THERAPIES, july/aug 2001, VOL. 7, NO. 4 Delivery Models for Acupuncture, Massage, Naturopathy