Process of requests submission for ranibizumab funding
We phrased an approval application form for the request of ranibizumab treatment funding (Figure 1). This application was submitted to the ophthalmic specialists team by the treating ophthalmologists employed at any of the 20 public hospital-affiliated retina units in the country, that treat individuals insured by Clalit Mushlam. Although usually not needed, supplementary medical records about the patients were sent directly from the retina units or the primary clinic in the community.
When all the required data were submitted, the form was forwarded to be evaluated by the ophthalmic specialists team by its representatives. This team used the information supplied on the application forms to assess the effectiveness of the treatment so far, as well as the need to start, continue, or renew ranibizumab administration. Notably, neither the ophthalmic specialists team nor the treating ophthalmologists requesting the funding, were provided with information regarding the program’s financial considerations of the steering committee and the expenses of by Clalit Mushlam. The patient, primary care unit personnel, and retina unit personnel are all blinded to the identity of the ophthalmic specialists that examined the application. The responses to the applications were sent via email, along with explanations, if necessary, to the patient's primary care physician and to the retina unit treating ophthalmologist who submitted the application form (Figure 2). A single funding approval covered 3 injection administrations and expired after 6-months period. The co-payment for patients prescribed ranibizumab, determined according to the regulations of Clalit Mushlam, was about $100 per injection (The drug and its administration's costs).
In patients that funding for the treatment with ranibizumab was denied, the treating ophthalmologists could either continue with bevacizumab injections, stop treatment or send an amended application form for re-evaluation of the request, if additional data were available (Figure 2).
Yield of the ranibizumab approval process
According to an external report supplied by Clalit Health Services prior to the study period, about 22,000 individuals were treated for retinal diseases with bevacizumab in 2011. By their estimation, they received about 200,000 injections annually in total.
Approximately 18,000 bevacizumab-treated patients (82%) had supplementary Clalit Mushlam insurance, accounting for 164,000 bevacizumab injections. The steering committee estimated the consumption of ~16,000 vials of ranibizumab annually, based on: (i) the numbers of Clalit Mushlam patients treated with bevacizumab, (ii) expected compliance of treating ophthalmologists and ophthalmic experts to the guidelines, (iii) adherence of patients to the treating ophthalmologists' recommendations, (iv) accepted treatment protocols for ranibizumab by each diagnosis, and (v) expected learning curve of the funding process by the treating ophthalmologists.
After the establishment of requests submission process for ranibizumab funding, Clalit Mushlam announced of it to the managers of all retina units in all public hospitals in Israel. Our study included all 5,642 patients (31% of bevacizumab-treated members of Clalit Mushlam), that applied for second-line ranibizumab treatment by a treating ophthalmologist in a public hospital-affiliated retina unit in Israel.
These patients included 2,867 females (51%) and 2,755 males (49%) patients of mean age 75.3 (SD = 11.13) years (Table 1). A total of 16,778 applications for the initiation or continuation of ranibizumab treatment in these patients were submitted for evaluation to the ophthalmic specialists team. The rate of application submission was 2.97 applications per patient. The requests submission rate was few requests at the beginning of the study, elevated up to 528 requests per month and its average was 357 requests per month throughout the study period (Figure 3). The entire process, from submission of the application form to either approval or rejection of the request, was completed within 72 hours.
Requests analysis and approval rates
Analysis revealed high approval rates of the funding requests. Throughout the process, as many as 80.2% of requests were approved after first application form was submitted (Table 2). Specifically, 84% of the first-time requests and 76% of continuation requests for ranibizumab treatment were approved after first application form. Moreover, even after rejection, 14% of all requests were approved after resubmissions of amended application form. Eventually, solely 808 requests (5.7% out of all requests) were rejected. Out of these, 288 requests (5.1% out of first-time requests) were rejected, thus did not receive any funding of the treatment, while 520 continuation requests were denied, thus these patients' treatment funding was terminated. Ultimately, 13,278 requests for subsidization were approved (average 2.35 approved requests per patient).
De facto, during the study period nearly 99.8% of approved requests received the series of three injections. According to a report of Clalit Health Services, as many as 41,739 injections were supplied to the retina units at any of the 20 public hospitals in Israel.
A diagnosis-based analysis reveals that the most frequent indication for treatment was AMD (61%), followed by DME (24%) and the rest had a complex diagnosis, CRVO/BRVO or miscellaneous (Table 1). The approval rate for requests differed significantly by indication (P < 0.001): 85%, 88%, 87%, 0% for AMD, DME, CRVO/BRVO and miscellaneous diagnoses, respectively. We further excluded patients with miscellaneous diagnoses, due to inadequacy to relevant indications. However, the likelihood of approval was lower for patients with AMD (OR 0.91, 95% CI 0.88-0.93), than for patients with DME (OR 1.27, 95% CI 1.77-1.37) or CRVO (OR 1.39, 95% CI 1.20-1.59), when compared to the rest of the diagnoses.
Overall, this process enabled treating about 18,000 patients with retinal disease effectively. While most of the patients received the less expensive, but not less effective treatment, only a reduced number of nonresponsders were determined as appropriate for funding with almost 10 times costlier treatment.