Appendix H
Summary of Cost Estimation Methodology for Outpatient Nutrition Therapy
SCENARIO ASSUMPTIONS
Baseline Utilization Scenario
Utilization in initial year and subsequent years to be consistent with selected previously published research and expert clinical judgment.
For example, Sheils and coworkers (1999) reported that from 1991 to 1996, the average utilization patterns for patients with diabetes reflected 464 nutrition sessions per 1,000 patient years. Using the estimated number of nutrition therapy visits per a 5-year period (seven for diabetes only, eight for diabetes with one other diagnosis, and nine for diabetes with two other diagnoses) and the NHANES III distribution between these categories (NCHS, 1997), the average number of visits per year is 1.64 per patient with diabetes. The actual rates provided by the two studies were 0.464 (28 percent utilization) and 0.193 (12 percent utilization). Thus, the baseline scenario was based on assumed utilization rate of 21 percent. This process was completed for each diagnosis.
Low Utilization Scenario
Utilization in initial year and subsequent years during the initial 5-year period may be lower that expected due to time needed to adjust healthcare system and physician referral patterns to new benefit. Using
the diabetes example from above, the low utilization cost estimate used 12 percent as the assumed utilization rate.
High Utilization Scenario
Utilization in initial year and subsequent years is projected to be higher that any previous studies to account for possible financial incentives in new benefit and potential impact of new practice protocols that were not being consistently followed during time of data collection of previous studies. For example, the research leading to the recommended number of diabetic visits was published in 1995 and the current protocols were published in 1998. In addition, some nutrition is assumed to be included in the new diabetes self management benefit, so the full initial nutrition therapy may not be warranted for persons using this benefit. For example, the high utilization scenario used 30 percent as the assumed utilization rate.
COST ESTIMATE ASSUMPTIONS
-
Conservative cost estimates (highest reasonable cost) were used.
-
Dollars are expressed in nominal terms and not discounted.
-
Medicare costs were provided by The Lewin Group, Inc. based on 1998 data from Office of the Actuary, Health Care Financing Administration.
-
Both Medicare costs and projected Medicare projected reimbursement rates were assumed to grow at the rate of 3 percent per year.
-
Reimbursement rates for nutrition therapy were assumed to be for individual sessions using rates established for diabetes care ($55.41 per session).
-
Estimated Medicare reimbursement costs were computed by estimating total direct costs and subtracting 20 percent cost sharing.
-
Future premium increases were estimated to be 25 percent of projected Medicare reimbursements.
-
Even though some Medicare-eligible would have other insurance, these cost estimates assumed that Medicare would be the payor for all nutrition therapy sessions received by beneficiaries.
-
Maximum number of nutrition therapy sessions were based on expert opinion and accepted protocols. They assume more sessions in year of initial diagnosis/referral than in subsequent years for most diagnoses. Even though most protocols are written for single diagnosis, patients usually present with combinations of diagnoses. Initial year treatments were estimated using highest number of sessions with an additional one visit per additional nutrition-related diagnoses.
-
Assumed that all beneficiaries receiving nutrition therapy received the maximum estimated number of nutrition therapy visits.
-
Assumed that all current Medicare beneficiaries would have disease prevalence similar to NHANES III data and that they would all become eligible for “newly diagnosed” benefit in the first year of coverage.
-
Assumed that Medicare beneficiary population changed at constant rate over the 5-year period. The estimated number of new beneficiaries was based on estimates from Office of the Actuary, Health Care Financing Administration.
-
All new beneficiaries were assumed to be treated as “initial year of diagnoses” for purposes of nutrition therapy. This assumes that previous health plan coverage did not offer nutrition as a covered benefit. (However some plans do; see discussion in chapter 14.)
ECONOMIC BENEFIT ASSUMPTIONS
Similar cost assumptions were used to quantify some potential economic benefits of nutrition therapy. Plus the following assumptions.
-
Nutrition therapy is usually provided in a multidisciplinary and multi-modality treatment plan. Therefore clinical outcomes of nutrition alone are difficult if not impossible to quantify accurately. Expert opinion and research were used to estimate amount of benefit that could be reasonably attributed to nutrition therapy. For example, diabetes trials indicate that intensive therapy including nutrition led to 0.9 percent point reduction in HbA1c, (e.g., 8.0 percent to 7.1 percent) however committee estimates for benefits were based on the assumption that 25 percent of this reduction might be reasonably attributed to nutrition therapy.
-
Clinical trial results were adjusted downward since maximum benefits reported are not likely to be replicated when nutrition therapy is provided to a broader population in a less controlled manner. For example, clinical trials with intensive nutrition therapy yielded a 6 percent reduction in cholesterol, while these estimates of benefits were based on a 3 percent reduction.
-
Economic benefits are likely for other diagnoses, however research showing quantifiable link between nutrition therapy and outcomes was not sufficient to prepare an estimate (e.g., for heart failure).
-
Various methodologies were used to prepare estimates for each diagnosis, and it is not appropriate to simply add the estimates since they overlap and accommodate persons with multiple diagnoses.
-
Economic benefits identified are not to be used as expected offsets for budget estimates.
-
Some benefits are likely not to occur immediately (i.e., within the first 2 years after therapy is initiated) and may well occur after the 5-year period of these estimates.
Estimated Reimbursement Cost in Billions to Medicare Summed over 2000 to 2004 (after adjustment for copayment and premium increase)
Baseline |
Low |
High |
$1.069 billion |
$740 million |
$1.97 billion |
Estimated Potential Benefit in Millions to Medicare Summed over 2000 to 2004
|
Baseline |
Low |
High |
Diabetes |
231 |
132 |
330 |
Hypertension |
83 |
52 |
167 |
Dyslipidemia |
89 |
54 |
154 |
Data were not available to permit estimating benefits of nutrition therapy for heart failure, pre-dialysis renal, and the one visit estimated for all remaining beneficiaries (over 65 years old without diagnoses and all beneficiaries under 65 years old) that were assumed to have other diagnoses that also could warrant referral for nutrition therapy.
REFERENCES
NCHS (National Center for Health Statistics). 1997. Third National Health and Nutrition Examination Survey (Series 11, No. 1, SETS version 1.22a). [CD-ROM]. Washington, D.C.: U.S. Government Printing Office.
Sheils JF, Rubin R, Stapleton DC. 1999. The estimated costs and savings of medical nutrition therapy: The Medicare population. J Am Diet Assoc 99:428–435.