Purpose
To evaluate whether cost, insurance prior authorization, pregnancy considerations, and subspecialty practice lead to changes in therapeutic decisions among uveitis and retina specialists in the treatment of intermediate and posterior uveitis.
Design
Prospective cross-sectional study.
Methods
A total of 934 uveitis and retina specialists across the United States were surveyed via e-mail regarding their choice in long-term therapy for 3 hypothetical patients with uveitic conditions (Behçet disease, birdshot retinochoroiditis, and intermediate uveitis). Respondents were asked to select first- and second-choice therapies and then reselect first and second choices assuming cost of all options was equal to the patient and there would be no issues with insurance prior authorization. In 1 case, they were additionally asked for their treatment preferences if the patient desired a pregnancy. Outcomes of interest were differences in therapy choice based on cost/prior authorization, pregnancy, and subspecialty practice.
Results
One hundred and six respondents (11.3%) completed the survey; 44 were uveitis specialists and 62 retina specialists. Cost and prior authorization affected the therapy choices of uveitis specialists treating ocular Behçet disease ( P = .008). Uveitis specialists and retina specialists differed in their first choice in therapy for each vignette when cost and prior authorization were equalized ( P = .0018, P = .0049, and P = .0156). Both uveitis specialists and retina specialists changed their therapeutic choices for intermediate uveitis when pregnancy was a consideration ( P = .00001 for uveitis specialists, P = .0044 for retina specialists).
Conclusions
Physician decision making in intermediate and posterior uveitis is affected by cost and prior authorization concerns, pregnancy considerations, and subspecialty practice.
The factors that affect therapeutic decisions by physicians have not been fully elucidated. Physicians are held to a professional and legal standard of care, which is the level and type of care that a reasonably competent and skilled health care professional with a similar background would have provided under the circumstances. A concerted push has been made across nearly all fields of medicine to incorporate more high-quality research evidence regarding diagnostic and therapeutic modalities into the management of patients to improve patient care. The goal of evidence-based medicine was originally conceptualized as an effort to apply lessons learned from accumulated basic science and clinical research studies to the treatment of individual patients. However, this model neglected a handful of factors that are also important in patient care. Clinical expertise and patient values and preferences are now recognized as important adjuncts in evidence-based decision making. Yet provider decisions may be affected by other influences, such as costs to the patient, insurance considerations, and subspecialty practice. The degree to which the above factors impact provider decision making within an evidence-based framework is still not well described.
The management of intermediate and posterior uveitis can serve as a case study in the analysis of provider decision making in uncommon diseases where extensive randomized controlled trials are not available. Systemic corticosteroids have long been the foundation for treating uveitis. However, ophthalmologists treating chronic uveitis attempt to choose therapies that will spare the patient from prolonged exposure to systemic steroids. Some of the long-term treatment options include local steroid implants, systemic immunomodulatory/immunosuppressive agents, and systemic biologic agents.
Although expert guidelines and research evidence give providers a framework for treating uveitis, we hypothesized that cost and prior authorization considerations would have an impact on physician decision making when choosing therapy for the treatment of uveitis. Furthermore, we hypothesized that uveitis specialists and retina specialists would differ in their approach to the treatment of uveitis. Finally, we hypothesized that pregnancy considerations would also cause significant alterations in physician decision making.
Methods
We developed an e-mail questionnaire with brief case vignettes of uveitic conditions. Exempt status approval from the Colorado Multiple Institutional Review Board was obtained, and all data were collected anonymously. Responses were collected using the online Survey Monkey tool. Respondents to the survey were asked whether they were members of the American Uveitis Society, the Macula Society, the Retina Society, the American Society of Retinal Specialists, or none of the above. Respondents were then presented with 3 clinical case vignettes ( Appendix ; Supplemental Material available at AJO.com ): a 45-year-old man with bilateral ocular Behçet syndrome and with systemic involvement; a 50-year-old man with bilateral birdshot retinochoroiditis; and a 25-year-old woman with bilateral pars planitis (intermediate uveitis). All patients had had an unsatisfactory response to either oral corticosteroids (Behçet and birdshot) or periocular triamcinolone acetonide (intermediate uveitis). Survey respondents were then asked whether they would manage the patient or refer to another specialist. Those who chose to manage the patient were asked for their first and second choice of treatment options. They were then asked for first and second treatment choices assuming that cost to the patient was equal for all medications and there would be no issues with prior authorization from the patient’s insurance provider (cost/authorization equalized). In addition, the respondents were asked what their choices would be for the patient with intermediate uveitis if she wished to become pregnant in the next 2 years. Possible treatment choices were dexamethasone implant, fluocinolone acetonide implant, methotrexate, cyclosporine, mycophenolate mofetil, infliximab, adalimumab, rituximab, and “other.”
The survey was sent to members of the American Uveitis Society discussion group as well as the Macula Society and American Society of Retina Specialists via publicly available e-mail lists. Respondents to the survey from the American Uveitis Society discussion group were categorized as uveitis specialists. Added to this group was anyone from the macula and retina e-mail groups who claimed membership in the American Uveitis Society. All other respondents from the macula and retina e-mail groups were categorized as retina specialists. The Survey Monkey tool allowed the survey to be completed only once per computer, preventing multiple responses by the same respondent.
For the analysis, dexamethasone and fluocinolone acetonide were grouped together as “local implants”; methotrexate, cyclosporine, and mycophenolate mofetil as “immunomodulators”; infliximab, adalimumab, and rituximab as “biologics”; and finally there was a grouping for “other.” An annual price for continuous therapy for each of the 8 options was calculated in a manner similar to that outlined by the Multicenter Uveitis Steroid Treatment (MUST) trial research group. The base cost for each medication was its average wholesale price per unit (AWP) as listed in RED BOOK Online, a pharmaceutical database that contains national drug cost and pricing information. The AWP was then multiplied by the number of units used over a year. For dexamethasone, infliximab, and rituximab, the appropriate Medicare physician reimbursement was added for in-office administration. For fluocinolone acetonide, the 3-year cost presented by the MUST group was averaged for 1 year to provide a more accurate representation of its annual cost.
As noted above, the responses from each of the treatment choice questions were pooled into 4 categories. Two-by-four contingency tables were constructed to compare the frequency of treatment choices between different questions and between the 2 subspecialties. Fisher exact tests were then performed on these contingency tables to detect differences in the frequencies (χ 2 tests were not used owing to small or zero expected values in some cells). The comparisons analyzed were: (1) uveitis specialists’ first and second choices before and after cost and prior authorization were equalized; (2) retina specialists’ first and second choices before and after cost and prior authorization concerns were equalized; (3) uveitis specialists’ first and second treatment choices before and after pregnancy is considered (Case 3 only); (4) retina specialists’ first- and second-choice treatments before and after pregnancy is considered (Case 3 only); and (5) between uveitis specialists and retina specialists for each question.
Results
A total of 106 out of 934 physicians returned the survey (response rate 11.3%). Forty-one of 192 (20.9%) responded from the American Uveitis Society discussion board, 24 of 300 (8%) from the Macula Society e-mail list, and 41 of 442 (9.3%) from the e-mail list of the American Society of Retina Specialists. Of the 65 respondents from the macula and retina lists, 3 claimed membership in the American Uveitis Society and were therefore included as uveitis specialists, leaving a total of 44 physicians categorized as uveitis specialists and 62 categorized as retina specialists.
A list of the treatment options available and their annual cost of continuous treatment is presented in the Table .
| Treatment Option | Annual Cost a |
|---|---|
| Dexamethasone b | $7932 |
| Fluocinolone acetonide c | $23 433 |
| Methotrexate d | $1248 |
| Cyclosporine e | $7300 |
| Mycophenolate mofetil f | $11 680 |
| Infliximab g | $22 528 |
| Adalimumab h | $21 060 |
| Rituximab i | $17 354 |
a Representative national cost approximations based on protocol utilized by the Multicenter Uveitis Steroid Treatment trial group and average wholesale price (and associated procedure and facility fees, where appropriate) from RED BOOK Online pharmaceutical database and the Centers for Medicare and Medicaid Services Physician Fee Schedule.
b 0.7 mg implant, bilaterally, 2.4 times per year (5 month duration in active disease).
c 0.59 mg implant bilaterally.
g 5 mg/kg intravenously, 50 kg patient, every 6 weeks.
h 40 mg/0.8 mL subcutaneous injection every 2 weeks.
Case 1: Behçet Disease
The distribution of first-choice therapy choices for ocular Behçet disease among uveitis and retina specialists can be seen in Figure 1 . Among the uveitis specialists, 39 out of 44 (88.6%) chose to personally recommend a treatment option for Behçet disease. Before equalization of cost/authorization, 0% (0/39) of uveitis specialists chose a local implant as their first-choice therapy, 43.6% (17/39) chose an immunomodulator, 46.2% (18/39) a biologic agent, and 10.3% (4/39) other. After equalization of cost/authorization, 2.6% (1/39) chose a local implant first, 7.7% (3/39) an immunomodulator, 79.5% (31/39) a biologic agent, and 10.3% (4/39) other ( P = .0008). For second-choice therapies before equalization of cost/authorization, 2.6% (1/39) of uveitis specialists chose a local implant, 25.6% (10/39) an immunomodulator, 59% (23/39) a biologic agent, and 12.8% (5/39) other. After equalization of cost/authorization, 0% (0/39) chose a local implant as their second-choice therapy, 15.4% (6/39) chose an immunomodulator, 74.4% (29/39) a biologic agent, and 10.3% (4/39) other ( P = .4158).
For the retina specialists 25 of 62 (40.3%) said they would treat the patient and not refer. Before equalization of cost/authorization, 12% (3/25) of retina specialists chose a local implant as their first-choice treatment option, 56% (14/25) chose an immunomodulator, 24% (6/25) a biologic agent, and 8% (2/25) other. After equalization of cost/authorization, 12% (3/25) again chose a local implant as their first choice, 40% (10/25) an immunomodulator, 40% (10/25) a biologic agent, and 8% (2/25) other ( P = .6781). For the second-choice therapies before equalization of cost/authorization, 16% (4/25) of retina specialists chose a local implant, 36% (9/25) an immunomodulator, 40% (10/25) a biologic agent, and 8% (2/25) other. After cost/authorization equalization, 20% (5/25) chose a local implant, 24% (6/25) an immunomodulator, 48% (12/25) a biologic agent, and 8% (2/25) other ( P = .8959).
The difference in first-choice therapy options between uveitis and retina specialists before equalization of cost/authorization was not significant ( P = .0599). After equalization of cost/authorization, the difference in first-choice options was significant between the 2 subspecialties ( P = .0018). The difference in second-choice therapies between uveitis and retina specialists was not significant before equalization of cost/authorization ( P = .1419), but was significant after equalization ( P = .0142).
Case 2: Birdshot Retinochoroiditis
The distribution of first-choice therapy choices for birdshot retinochoroiditis among uveitis and retina specialists can be seen in Figure 2 . There were 39 out of 44 uveitis specialists (88.6%) who said they would treat the patient with birdshot retinochoroiditis. Before equalization of cost/authorization, 10.3% (4/39) of uveitis specialists chose a local implant as their first-choice therapy, 84.6% (33/39) chose an immunomodulator, 0% (0/39) a biologic agent, and 5.1% (2/39) other. After equalization of cost/authorization, 10.3% (4/39) chose a local implant first, 76.9% (30/39) an immunomodulator, 10.3% (4/39) a biologic agent, and 2.6% (1/39) other ( P = .2193). For second-choice therapies before equalization of cost/authorization, 15.4% (6/39) of uveitis specialists chose a local implant, 53.8% (21/39) an immunomodulator, 23.1% (9/39) a biologic agent, and 7.7% (3/39) other. After equalization of cost/authorization, 15.4% (6/39) chose a local implant as their second-choice therapy, 46.1% (17/39) chose an immunomodulator, 33.3% (13/39) a biologic agent, and 7.7% (3/39) other ( P = .7988).
Among the retina specialists, 30 of 62 (48.4%) completed the questions regarding treatment options. Before equalization of cost/authorization, 30% (9/30) of retina specialists chose a local implant as their first-choice treatment option, 56.7% (17/30) chose an immunomodulator, 0% (0/30) a biologic agent, and 13.3% (4/30) other. After equalization of cost/authorization, 36.7% (11/30) chose a local implant as their first choice, 46.7% (14/30) an immunomodulator, 3.3% (1/30) a biologic agent, and 13.3% (4/30) other ( P = .8745). For the second-choice therapies before equalization of cost/authorization, 20% (6/30) of retina specialists chose a local implant, 50% (15/30) an immunomodulator, 16.7% (5/30) a biologic agent, and 13.3% (4/30) other. After cost/authorization equalization, 26.7% (8/30) chose a local implant, 46.7% (14/30) an immunomodulator, 13.3% (4/30) a biologic agent, and 13.3% (4/30) other ( P = .9602).
The difference in first-choice therapy options between uveitis and retina specialists before equalization of cost/authorization was significant ( P = .0314). The difference in first-choice options was also significant between the 2 subspecialties after equalization of cost/authorization ( P = .0049). The difference in second-choice therapies between uveitis and retina specialists was not significant before equalization of cost/authorization ( P = .7611) or after equalization ( P = .2130).
Case 3: Pars Planitis (Intermediate Uveitis)
The distribution of first-choice therapy choices for pars planitis among uveitis and retina specialists can be seen in Figure 3 and Figure 4 . Figure 3 demonstrates treatment selection before and after equalization of cost/authorization, while Figure 4 shows treatment selection before and after pregnancy is considered. Among the uveitis specialists, 39 of 44 (88.6%) said they would personally treat the patient. Before equalization of cost/authorization, 30.7% (12/39) of uveitis specialists chose a local implant as their first-choice therapy, 51.3% (20/39) chose an immunomodulator, 5.1% (2/39) a biologic agent, and 12.8% (5/39) other. After equalization of cost/authorization, 25.6% (10/39) chose a local implant first, 51.3% (20/39) an immunomodulator, 10.3% (4/39) a biologic agent, and 12.8% (5/39) other ( P = .8687). After consideration of a possible pregnancy within 2 years, 59% (23/39) of uveitis specialists made a local implant their first choice, 2.6% (1/39) chose an immunomodulator, 12.8% (5/39) a biologic agent, and 25.6% (10/39) other ( P = .00001, compared with first choice before equalization of cost/authorization). For second-choice therapies before equalization of cost/authorization, 20.5% (8/39) of uveitis specialists chose a local implant, 46.1% (18/39) an immunomodulator, 20.5% (8/39) a biologic agent, and 12.8% (5/39) other. After equalization of cost/authorization, 20.5% (8/39) chose a local implant as their second-choice therapy, 41% (16/39) chose an immunomodulator, 25.6% (10/39) a biologic agent, and 12.8% (5/39) other ( P = .9864). After considering a possible pregnancy in the next 2 years, 51.3% (20/39) of uveitis specialists chose a local implant as second choice, 0% (0/39) an immunomodulator, 15.4% (6/39) a biologic agent, and 33.3% (13/39) other ( P = .00001, compared with second choice before equalization of cost/authorization).
