The advent of
intravitreal drug therapy for retinal vascular disease over the past decade constitutes a therapeutic revolution of unprecedented scale in the history of ophthalmology. The ability to minimize loss of vision in more than 90 percent of patients with new-onset neovascular age-related macular degeneration and to improve vision in approximately 40 percent of patients over a period of years has resulted in untold clinical benefit.1
More recently, a similar treatment benefit has been demonstrated in retinal venous occlusive disease and diabetic macular edema.2-4
This report will discuss the epidemiology of intravitreal injections—examining the past growth, estimating the current number, and speculating on the future growth of intravitreal injections with a consideration of potential health policy implications.
The best available data on the epidemiology of intravitreal injections comes from the AMA/Specialty Society RVS Update Committee Database.5
However, this database is limited to a 5-percent sample of the Medicare Physician Fee for Service schedule. It does not include Medicare Advantage or non-Medicare utilization and therefore is not a comprehensive source.
Prior to 2001, the number of intravitreal injections (CPT code 67028) in the RUC database was relatively stable at approximately 4,500 per year, primarily for the treatment of end-ophthalmitis. In 2002, the number of injections tripled to 15,000 (largely triamcinolone) and by 2004 it was 83,000. The breakthrough year was 2005 with growth to 252,000 following the first reports on pegaptanib, bevacizumab and ranibizumab. By 2008, more than 1,000,000 injections were performed. In 2011, aflibercept became available and total intravitreal injections topped 2,000,000. In 2012, 2,354,753 were reported in the RUC database.
Although precise numbers are not available for total 2013 utilization, a reasonable estimate can be calculated by using a conservative growth rate of 10 percent (substantially less than the growth rates of 14 to 27 percent over the period 2008 to 2012), resulting in approximately 2.4 million injections. A total number of injections for Medicare can be estimated by including Medicare Advantage beneficiaries at ~30 percent of the Medicare population for an approximate 3.1 million injections in the total Medicare program. Although Medicare-eligible patients continue to constitute the majority of individuals treated with intravitreal injections, an increasing number of younger patients are being treated for diabetic macular edema and retinal venous occlusions. Based on utilization data from my group, I estimate that approximately 30 percent of patients receiving intravitreal injections are not in Medicare. This results in an estimated total of 4.1 million injections in 2013.
Tough to Predict
Now that we have established a baseline, what can we predict about future growth? Although I am a big fan of the Yogi Berra aphorism that, “Its hard to make predictions, especially about the future,” I think we can use our knowledge of demographics and the likely timeline for future disruptive technologies to estimate what to expect over the next three years. Using a conservative yearly growth rate of 10 percent suggests approximately 5.9 million injections in 2016. If we assume a less conservative, but still realistic growth rate of 20 percent the estimate is 7.9 million. The logistics of delivering these numbers of injections will be challenging.
The RUC database reveals interesting demographics about Medicare beneficiaries receiving intravitreal injections. As expected, there is substantial racial disparity: 93.9 percent white, 2.5 percent black, 1.2 percent Hispanic and 2.4 percent other. A gender disparity also exists, with 67 percent female and 33 percent male. A remarkable 34 percent are greater than 85 years of age and 77 percent are greater than 75 years of age. More than 97 percent are age-eligible for Medicare, with 2 percent disabled and 1 percent with end-stage renal disease.
An analysis of the indications for intravitreal injections based on ICD-9 coding reveals that 96.5 percent are in the ICD 362 family that includes most retinal vascular diseases. Unfortunately, the database does not have granularity within the ICD 362 family to determine the relative proportion of neovascular AMD, diabetic retinopathy or retinal vascular occlusions. Diabetes, coded as ICD 250, occurs in 2.5 percent with glaucoma (ICD 365) at 0.2 percent and other CPT diagnoses at 0.4 percent constituting the remainder.
The logarithmic growth of intravitreal injections is attracting increasing health policy scrutiny. Despite consensus on the clinical efficacy and safety of intravitreal injections, the rapidly expanding costs are a concern. These costs include the examinations and retinal imaging associated with intravitreal injections, and of course the cost of the drugs. The reimbursement for the intravitreal injection procedure was substantially decreased by Medicare in 2010 due to a shortened procedure time. Although the continuing volume growth may result in additional review, I do not see how ophthalmologists can further significantly cut the time necessary to deliver this essential treatment. Furthermore, cuts in physician reimbursement will have minimal effect on overall costs.
In June of 2013, the General Accounting Office released a report on Part B drug costs for 2010.6 In this report, ranibizumab was the third highest-expenditure Part B drug for Medicare beneficiaries at $1.18 billion. In 2010, Medicare expenditure for intravitreal injections was approximately $200 million. Even with the addition of other related physician services such as ocular examinations and imaging, the primary driver of expenditures associated with intravitreal injections is the cost of the drugs. New indications such as diabetic macular edema, central vein occlusion and branch vein occlusion as well as the advent of aflibercept and ranibizumab for DME have substantially further increased drug expenditures. In 2013, U.S. sales of ranibizumab and aflibercept were approximately $1.73 billion and $1.35 billion respectively, according to available financial reports. Therefore, in composite, drugs for intravitreal injections are now likely the highest expenditures for Part B drugs.
The health policy implications of the increasing cost of intravitreal therapies are becoming apparent. Some payers are now requiring step therapy protocols that mandate the initial use of bevacizumab. Ranibizumab or aflibercept may only be used in cases of treatment failure. The criteria for treatment failure are often poorly defined. The mandated use of bevacizumab has been complicated by the recent Drug Quality and Security Act that has created access and logistical issues for use of off-label bevacizumab in several parts of the country.
The ACA Impact
The Affordable Care Act is also likely to impact intravitreal injections, particularly the choice of drug. Starting in 2015, a value-based modifier will be a factor in Medicare physician payment. The details of VBM are uncertain, but it is likely that the use of expensive drugs, regardless of clinical indication, will be associated with a financial penalty as part of a quality and cost assessment. This penalty will be a 2-percent cut on all Medicare physician payments. Considering the limited margin on Part B drugs, some ophthalmologists may be unable to continue using expensive drugs. (And as noted above, federal and state compounding regulations may make it increasingly unattractive to use less-expensive drugs).
Patients will also be affected. The new health exchange plans and many commercial health insurance plans are characterized by high deductible expenses that may preclude the use of expensive drugs. The exploding cost of intravitreal therapy will present increasingly complex challenges to ophthalmologists, payers and most importantly, patients.
Dr. Williams practices at Associated Retinal Consultants, William Beaumont Hospital, 3535 West 13 Mile Rd. #344, Royal Oak, MI 48073. Contact him at
1. Comparison of Age-related Macular Degeneration Treatments Trials (CATT) Research Group. Ranibizumab and bevacizumab for treatment of neo-vascular age-related macular degeneration: Two year results. Ophthalmology 2012;119;1388-98.
2. RESTORE Extension Study Group. Two-year safety and efficacy of ranibizumab 0.5 mg in diabetic macular edema: Interim analysis of the RESTORE extension study. Ophthalmology 2013;120:2004-12.
3. Sustained benefits from ranibizumab for macular edema following central retinal vein occlusion: Twelve-month outcomes of a phase III study. Ophthalmology 2011;118:204-9.
4. GALILEO Study Group. Intravitreal Aflibercept Injection for Macular Edema Resulting from Central Retinal Vein Occlusion: One-Year Results for the Phase 3 GALILEO Study. Ophthalmology 2014;121:202-8.
5. RBRVS DataManager Online. Copyright American Medical Association
6. James Cosgrove. United States Government Accountability Office. GAO-13-739T. Medicare Information on Highest-Expenditure Part B Drugs. June 28, 2013.