By David Kiefer, MD, Editor
Synopsis: Acupuncture improved symptoms of seasonal allergic rhinitis, but it is expensive.
Source: Reinhold T, et al. Cost-effectiveness for acupuncture in seasonal allergic rhinitis: Economic results of the ACUSAR trial. Ann Allergy Asthma Immunol 2013;111:56-63.
The researchers of this randomized, controlled trial attempted to determine the clinical and cost effectiveness of acupuncture for seasonal allergic rhinitis (SAR) sufferers in Germany. This study was an additional analysis of the ACUSAR (Acupuncture in Seasonal Allergic Rhinitis) trial in which study participants with at least 2 years of moderate-to-severe allergic rhinitis were randomized to 12 sessions of acupuncture plus antihistamine, 12 sessions of sham acupuncture plus antihistamine, or just antihistamine. With respect to the antihistamine, the participants were allowed to take up to two doses of cetirizine daily. There were numerous exclusion criteria as listed in Table 1.
|Table 1: Inclusion and Exclusion Criteria for the Acupuncture in SAR Study|
The treatments continued for 8 weeks, and then participants were followed for an additional 8 weeks during which the control group also received acupuncture. Quality of life (QOL) as assessed by the Medical Outcomes Study 36-Item Short Form Health Survey was measured at baseline, 8 weeks, and 16 weeks. The QOL score was converted to a "health utility" score that the researchers had used previously; a perfect QOL equates with a utility score of 1, whereas a 0 represents the absence of full QOL, or death. All costs associated with the treatment of the study participants’ SAR were calculated using estimates of either out-of-pocket acupuncture costs or third-party payer reimbursements (the former was labeled "society’s perspective," the latter "third-party payer’s perspective"), as well as patient questionnaires at baseline and weeks 8 and 16 about clinic visits and hospitalizations, days off of work (only included in "society’s perspective"), and medication costs. The researchers used a series of calculations to arrive at an Incremental Cost-Effectiveness Ratio (ICER), which was the cost for gaining one additional Quality-Adjusted-Life-Year (QALY). An ICER threshold of $67,575 (50,000 euros) per QALY gained was used, an acceptable level as compared to past international research.
A total of 422 participants were randomized to the three arms, and cost and QOL data were available for analysis for 364 participants. The baseline cost results, essentially the estimates of what was spent in the 8 weeks before randomization, for the three arms are listed in Table 2. The two acupuncture groups had more outpatient visits and more work lost than the control group, accounting for most of the observed differences at baseline. After the first 8 weeks of the study interventions, costs listed in the three study arms are also shown in Table 2, as were health utility scores, the QOL representation. No P values were given for the health utility scores, making it difficult to determine if any of these trends were statistically significant. However, from this table it does appear that the sham acupuncture group had the most costs (presumably from ineffectual treatment and more outpatient visits, more medication use, etc.), that the acupuncture group had the highest utility score, and that the control group showed almost no improvement in health utility.
|Table 2: Society's Cost (Using Out-of-pocket Acupuncture Cost and Days of Work Lost) and Third-party Payer's Cost at Baseline and After 8 Weeks for the Three Arms of the Research Trial|
|Acupuncture group||Sham acupuncture group||Antihistamine only group|
|Baseline||8 weeks||Basiline||8 weeks||Basline||8 weeks|
|Third-party payer's cost||$41.95||$333.83||$24.44||$336.96||$20.77||$21.04|
|Health utility score, from 0-1||0.750||0.798||0.744||0.770||0.762||0.768|
The researchers also computed costs and QOL results after 16 weeks of treatment, but pointed out that inter-group comparisons were complicated due to the control group also receiving acupuncture during this time. The 16-week mark was considered most interesting for an "observation" period to determine the persistence of the results from the original 8-week intervention. On that note, the utility health score at week 16 for the acupuncture group was 0.821, for the sham acupuncture group was 0.800, and for the control group was 0.832. Again, without P values, it is difficult to determine statistical significance, but one read of these data is that the acupuncture and sham acupuncture groups continued to improve during the 8-week "observation" period, while the control group, now receiving acupuncture, showed similar, if not higher, health utility scores, than the original acupuncture groups.
With respect to cost-effectiveness, from baseline to week 8, only acupuncture showed statistically significant higher QALY (P = 0.01); QALY for the sham acupuncture group did not change from baseline to week 8 (P = 0.40). However, both the acupuncture and sham acupuncture groups accrued more costs than the control group, as would be expected due to the costs of the treatments themselves; apparently even sham acupuncture costs money. A range of ICER values were calculated (see Table 3) for the two treatment groups vs the control group; as a reminder, the threshold per QALY gained was $67,575 in order to justify, from a cost-effectiveness standpoint, a particular intervention. Sham acupuncture, for all intents and purposes, did not meet the criteria for a justifiable intervention, whereas normal acupuncture, the researchers posit, had 1.3% probability (from the society’s perspective) and 22% probability (from the third-party payer’s perspective) of being "worth its money."
|Table 3: Incremental Cost-Effectiveness Ratio (ICER), Which is the Cost for Gaining One Additional Quality-Adjusted-Life-Year (QALY), for the Two Treatment Groups vs the Control Group|
|Acupuncture Group vs Control||Sham Acupuncture Group|
|Third-party payer's cost||$28,121-100,802||$66,441-344,022|
Two other points are interesting in this analysis. First, no one needed to be admitted to the hospital for SAR, not a surprise, but important to know in the cost calculations. And, secondly, the three arms of this study had varying visits to outpatient clinics because of SAR, the differences of which weren’t analyzed statistically in terms of significance and P values: 4.3% of patients in the acupuncture group went to a clinic because of their SAR vs 4.7% of the sham acupuncture group and 16.7% of the control group.
This was an ambitious research endeavor. It would have been a worthy effort had the researchers simply created a tested and testable research protocol for studying acupuncture; the creation of sham acupuncture points, where needles are used to pierce the skin at sites not associated with any meridian, is complex in and of itself. But then, in addition to this, the researchers created a complex methodology for calculating costs and cost-effectiveness, addressing the important question of whether or not the benefit received from this integrative therapy is worth the extra cost. For many of us, mixing economics with clinical care is walking a fine line between compromising adequate care on one hand vs overspending on the other. How many clinicians know, for instance, the cost of a rapid throat test for Streptococcus, and does that matter?
A few interesting conclusions can be drawn from this research study. The first is that both acupuncture and sham acupuncture are costly, but only correctly administered acupuncture passes a cost-effectiveness evaluation, though not 100% of the time (rather 1.3-22% of the time). Part of this makes sense, as we wouldn’t expect sham acupuncture to achieve therapeutic goals. However, even the acupuncture group wasn’t universally cost-effective, especially when factoring in society’s seasonal allergic rhinitis costs, including days off work and out-of-pocket acupuncture expenses. Perhaps the gentle nudge from acupuncture in this scenario can’t overcome such a significant cost burden in this research methodology. It makes one wonder, though, how these analyses would change if "intangibles," such as pscyho-social-mind-body benefits from the provider-patient encounter itself, were factored in. A hint to this effect exists in the health utility (QOL) scores: Even sham acupuncture showed a benefit in this scale. Sham acupuncture fell behind regular acupuncture in accrued costs, leading it to be cost-ineffective, but it may agree with other studies showing a clinical benefit beyond the acupuncture itself.1
There are some perplexing aspects to this study. A fairly large number of study participants (58 of 422) did not have cost or QOL data available, and there were no reasons given for these "dropouts" nor was there any indication whether they were included in an intention-to-treat analysis. Presumably they weren’t included in this way, compromising the results. Furthermore, the extensive exclusion criteria may leave clinicians scratching their heads, wondering how generalizable these results are to their patient population. And, of course, cost-effectiveness thresholds create controversy and disagreement about QALY: When, based on cost, do we deny a treatment for someone? It would have been helpful to see other calculations relevant to clinicians, such as the number-needed-to-treat.
Should clinicians now turn to acupuncture as a first-line therapy for people with SAR? Perhaps the answer is "Yes, if you can get a third-party payer to pick up the tab." This study is anything but the final word on the topic, but it is helpful to see a research group taking on the difficult task of blending clinical effectiveness and cost, using terms such as ICER and QALY that we will all surely be seeing more of as health care costs rise and the push for evidence-based medicine only intensifies.