Data Sources

We obtained Medicare standard analytic claim files from the US Centers for Medicare & Medicaid Services. Medicare is the primary health insurer for the US population 65 years and older. All beneficiaries receive Medicare Part A benefits, which include inpatient care, and over 90% enroll in Medicare Part B, which helps cover medically necessary outpatient services. We obtained longitudinal claim files and denominator files for all fee-for-service Medicare beneficiaries who had at least 1 claim with a diagnosis of neovascular AMD (International Classification of Diseases, Ninth Revision, Clinical Modification [ICD-9-CM] diagnosis code 362.52) between January 1, 2004, and December 31, 2010. For this analysis, we used Medicare Part B outpatient claims files, which contain claims from outpatient institutional providers, and carrier claims files, which contain claims from noninstitutional professional providers (eg, physicians), including claims for injectable medications. The denominator files include demographic characteristics, information about Medicare eligibility and enrollment, and death dates, if applicable. The institutional review board of the Duke University Health System approved this study.

Study Population

The study population included all eyes in which patients received an anti-VEGF injection for neovascular AMD between January 1, 2005, and December 31, 2010. To identify the study population, we first selected all outpatient and carrier claims for intravitreous injections (Current Procedural Terminology [CPT] code 67028) of bevacizumab or ranibizumab and a supporting diagnosis of neovascular AMD. To identify the medications associated with intravitreous injections, we first searched for medication-specific Healthcare Common Procedure Coding System (HCPCS) medication reimbursement codes. For periods during which medication-specific codes were not available, we identified the medications by using nonspecific reimbursement codes in combination with reimbursement amounts from the Medicare quarterly pricing reports ( Supplemental Table 1 , available at AJO.com ). Using modifier codes present on each injection claim, we determined which eye was treated. We considered the date of the earliest observed anti-VEGF treatment per eye to be the index anti-VEGF treatment date for that eye. We restricted the analysis to beneficiaries who were 65 years or older and living in the United States on the index date. To help ensure that we were observing the first anti-VEGF treatment associated with each eye, we also required that the index date follow a 1-year period of continuous fee-for-service Medicare enrollment without claims for anti-VEGF therapy.

Treatment and Discontinuation

We searched outpatient and carrier claims in the 2 years after the index date for procedures related to the management of neovascular AMD. Specifically, we searched for intravitreous injections of anti-VEGF agents, identified as described above, and optical coherence tomography (OCT) procedures (CPT code 92135). We defined anti-VEGF treatment as a claim for an intravitreous injection of an anti-VEGF agent. We defined a treatment visit as a claim for either anti-VEGF treatment or OCT. We defined discontinuation of anti-VEGF therapy as the absence of anti-VEGF treatment for at least 12 months, and discontinuation of AMD treatment visits as the absence of treatment visits for at least 12 months. We considered discontinuation of therapy to begin on the first day of the 12-month period.

Other Variables

To understand the demographic and clinical characteristics of patients in the study population at the time of treatment initiation, we used age, sex, and race information from the denominator files and indicators of ocular comorbid conditions—cataracts, glaucoma, diabetic macular edema, and diabetic retinopathy—for the index year. We defined cataracts and glaucoma using the Chronic Condition Warehouse indicators. We searched claims from the index year for diagnoses of diabetic macular edema (ICD-9-CM diagnosis code 362.07 or a combination of code 362.53 or 362.83 with code 250.xx on the same claim) and diabetic retinopathy (ICD-9-CM code 362.01 or 362.02) and for treatment with photodynamic theraphy (CPT code 67221 or 67225) or thermal laser photocoagulation (CPT codes 67220 or 0017T).

Statistical Analysis

We describe the baseline characteristics of patients in the study population at the time of the first anti-VEGF treatment by presenting frequencies with percentages for categorical variables and means with SDs for continuous variables. We present these characteristics by patient and by treated eye. Patients who initiated treatment in both eyes during the study period contributed data twice to the eye-specific analyses, corresponding to the timing of the initial injection in each eye.

We calculated the frequency of anti-VEGF treatments and treatment visits for neovascular AMD per treated eye in the first and second year after initial treatment using the cumulative incidence function. We considered both anti-VEGF treatments and treatment visits to be recurring events, enabling us to categorize patients into the following groups according to treatment intensity: none (0 injections or visits), low frequency (1–4 injections or visits), medium frequency (5–6 injections or visits), and high frequency (7 or more injections or visits). We calculated the mean number of treatments and treatment visits per eye using the mean frequency function. Both the cumulative incidence and mean frequency functions account for the competing risk of mortality and the fact that follow-up differs across patients as a result of data censoring. We defined the date of censoring as the earlier of December 31, 2010, or the date on which the patient died or enrolled in Medicare managed care, if applicable. Among patients with at least 12 months of data after initiation of anti-VEGF treatment (ie, those who initiated treatment before January 1, 2010), we estimated rates of discontinuation using Kaplan-Meier methods. We used SAS version 9.2 (SAS Institute Inc, Cary, North Carolina, USA).