MNT: Disorders of Lipid Metabolism (2015)

Citation:

Troyer JL, McAuley WJ, McCutcheon ME. Cost-effectiveness of medical nutrition therapy and therapeutically designed meals for older adults with cardiovascular disease. J Am Diet Assoc. 2010; 110: 1840-1851.

 

PubMed ID: 21111094
 
Study Design:
Cost-effectiveness study
Class:
M - Click here for explanation of classification scheme.
Quality Rating:
Positive POSITIVE: See Quality Criteria Checklist below.
Research Purpose:

To assess the cost-effectiveness of MNT and therapeutic meals for older adults with hyperlipidemia or hypertension.

Inclusion Criteria:
  • Aged 60 years or older
  • Diagnosed by a physician with hyperlipidemia or hypertension
  • Willing and able (as certified by their physician) to participate.
Exclusion Criteria:
  • Coronary artery bypass
  • Diabetes
  • Myocardial infarction within 12 months before enrollment
  • Renal insufficiency
  • Major surgery within 12 weeks before enrollment
  • Unstable angina
  • Uncompensated congestive heart failure
  • Liver disease
  • Pancreatic disease
  • Current serious gastrointestinal disease
  • Current cancer of other life-threatening diagnoses
  • Individuals with a very high knowledge of dietary and lifestyle management issues required by the diagnosis, based on a Heart Disease and Food Selection Diet/Lifestyle Knowledge Questionnaire. Participants who correctly answered 18 questions were excluded because of potential lack of benefit from MNT.
Description of Study Protocol:

Recruitment

Participants were recruited through referrals from physicians, clinics, agencies and churches, along with attendance at health fairs, flu clinics, housing for older adults and advertisements in various media sources.

Design

  • Cost-effectiveness study based on randomized, controlled, prospective trial
  • Randomization with minimization strategy was used to reduce differences in the stratification criteria across study arms and is described in detail in reference number 12.

Blinding Used

Not applicable because of nature of intervention.

 Intervention

  • Literature arm: Control arm; delivery by mail of commonly available literature on management of the participant's disease; telephone responses by RD to specific questions; no other interaction with RD until conclusion of the study
  • Therapeutic meal arm: Pre-intervention assessment but no nutrition counseling; participants received seven diagnosis-appropriate meals a week, delivered once per week (based on ADA MNT protocols and Administration on Aging Nutrition Program dietary requirements); study RDs occasionally contacted a participant to assure that the meals were being managed and consumed properly; meal consumption was assessed during an eight-month retention telephone call, in which participants were asked how many project meals they consumed in an average week; participants were encouraged to call in with questions or problems.
  • MNT arm: Participants received MNT from the study RDs, using ADA's Hyperlipidemia Medical Nutrition Therapy Protocol; for participants with hypertension, study RDs followed the hyperlipidemia protocol with appropriate adjustments for those diagnosed with hypertension. This was due to lack of a finalized MNT protocol for hypertension; in-home MNT sessions; intervention occurred over three sessions; the MNT intervention was intentionally individualized for each participant and independent expert consultant audited study RD records to determine the reliability and validity of MNT sessions.
  • MNT-plus-therapeutic meals arm: Participants received both of the previously prescribed interventions. 

Statistical Analysis

  • Approaches to cost calculation: Three cost categories were used; only intervention (and pre-intervention) costs of therapeutic meals and MNT; medical costs that were directly related to cardiovascular disease were added to intervention; medical costs indirectly related to cardiovascular disease were included. Cardiovascular-related visits were separated from non-cardiovascular-related visits.
  • Where the benefits were significantly different across groups, group costs were compared in light of averaged health and lifestyle benefits using costs-effectiveness analyses, following the approach described by Drummond (refs 30 and 31)
  • Incremental cost-effectiveness ratio and the net benefit approach served as measures of cost-effectiveness
  • Incremental cost-effectiveness ratio (ICER): Considers the added cost of the new intervention relative to the added health effect, where low ICERs imply lower costs per outcome than higher ICERs
  • The Literature arm served as the control group
  • Costs and quality-associated life years (QALYs) were not discounted due to the one-year time frame of the study.
  • The change in costs was estimated using a multivariate model. QALYs were modeled using ordinary least-squares regression, where the difference in outcome for the average participant attributable to the intervention was found by examining the coefficient on a binary indicator for the intervention. Costs were estimated using a generalized linear model with a gamma-distribution and log link.
  • Following the multivariate estimation processes, the ratio of the marginal effect of each intervention from the cost model to the coefficient on each the intervention from the QALYs model was the computed ICER
  • Using the ICER and a threshold that is the maximum acceptable willingness to pay for an outcome, an intervention is cost-effective if the ICER ratio is lower than this threshold
  • The ICER provides no guidance on whether added costs are worth any gains obtained, thus a net monetary benefit approach was also used. This allows for characterization of confidence intervals in the presence of multivariate estimation, controlling for baseline group differences.
  • The new intervention should be adopted if the net benefits exceed the net intervention costs
  • Net benefit is computed for each participant, given that participant's QALYs (E), costs (C) and a value for the maximum society is willing to pay for an additional QALY (threshold)
  • Net monetary benefit is then modeled (using ordinary least-squares) as a function of whether the individual received the intervention controlling for baseline variables that may differ across treatment arms. A confidence interval can then be constructed.
  • For each threshold, the probability that the intervention is cost-effective (a one-tailed test where the null hypothesis is that the intervention is not cost-effective) depends on whether the estimated coefficient on the binary indicator for the intervention is positive or negative. 
Data Collection Summary:

Timing of Measurements

  • Enrollment began in May 2003 and the study duration was 52 weeks
  • Baseline, six months (mid-intervention) and one year (post-intervention). 

Dependent Variables

  • QALYs (quality-adjusted life-years): Short-Form 36 Health Survey was the basis for constructing QALYs; a SF-6D utility index was calculated from six of the eight dimensions, with scores ranging from zero (as bad as death) to one (full health) and reflects how the general public values the reported health status; composite measure of quality of life, with more weight given to better health outcomes
  • Intervention cost measurement: Assessed from a societal perspective by obtaining data on all costs in three broad categories (MNT administration, therapeutic meal production delivery and participant-level medical costs). Costs associated with the research study were separated from costs associated with the actual delivery of therapeutic means and MNT. Study personnel kept detailed time and activity logs, and actual wage data were used to compute intervention labor costs. The intervention costs for the Literature group were set to zero due to the small cost of postage.
  • Health Care and Medication Cost measurement: Weekly health diaries were used, along with reminder calls and weekly reminder notes in the diaries; included sections on hospital use, nursing home use, emergency department visits, doctor visits and other health-related office visits, hours for informal help, professional nursing care, professional housekeeping, professional personal care, other professional in-home care. Prices and costs of these items were taken from a variety of sources described in the article; medication inventories were also used; costs from all sources were inflated to 2004 dollars using the Consumer Price Index.

Independent Variables

  • Literature arm (control)
  • Therapeutic meal arm
  • MNT arm
  • MNT-plus-therapeutic meal arm.

 

Description of Actual Data Sample:

Initial N

  • The original sample size was based on power tests clinical measures, which were not the focus of the current study
  • N=321 eligible people that were enrolled.

Attrition (Final N)

  • A total of 22 withdrew and did not permit further data collection
  • A total of 268 completed the study
  • A total of 31 ended the intervention but permitted data collection (intent-to-treat) [literature (zero); meals (14); MNT (4); meals and MNT (13)]
  • Analysis excludes four participants who died during the study and one observation with missing QALY data
  • A total of 298 participants were included in the final sample (16% male); for each group:
    • Control: N=7
    • Meals only: N=80
    • MNT only: N=73
    • Meals + MNT: N=70.

Age

Aged 60 years or older; no other information was given on age.

Ethnicity

A total of 62% were White.

Other Relevant Demographics

  • Approximately one-third were married (34%)
  • The average participant was taking slightly more than five prescription drugs (81% used a prescription drug for cardiovascular disease)
  • A total of 3% were homebound
  • Activities of daily living: 13.46 (SE=0.06)
  • Instrumental activities of daily living: 13.29 (SE=0.08)
  • Slightly more participants were diagnosed with hypertension than hyperlipidemia, more than half had both diagnoses.

Anthropometrics

  • Average BMI: 29.6kg/m2 (SE=0.37)
  • Average waist circumference: 94.7cm (SE=0.86).

Location

 Not specifically stated, but author affiliations imply the study was conducted in Charlotte, North Carolina.
Summary of Results:

Key Findings

  • Per-person cost: A total of $244 for an average of three sessions of home-delivered MNT; $1,565 ($4.30 per meal) for meals only; $1,877 for MNT+meals
  • For the Literature arm, the mean utility measurements indicated a decline in health during the year
  • For the intervention groups, the average QALYs were higher than the baseline measurement, indicating an improvement in health-related quality of life
  • QALYs for the Meals and MNT+Meals, but not for MNT only, were significantly higher than the Literature arm
  • MNT by itself was the most cost-effective intervention arm
  • MNT+Meals did not appear to be more cost-effective than each intervention on its own
  • All three interventions met established standards for cost-effectiveness
  • MNT only and Meals only attained higher probabilities of cost-effectiveness at lower willingness-to-pay levels when compared with MNT+Meals.

Other Findings

Thirty chart audits were completed for each study RD using the ADA Medicare MNT chart audit tool and the levels of conformance were greater than 95%.
Author Conclusion:
  • MNT and therapeutic meals was not more cost-effective than each intervention alone. The study results also support extending Medicare funding for MNT to individuals with hypertension and hyperlipidemia who do not also have diabetes.
  • Future research on lifestyle changes for hypertension and hyperlipidemia should include more individuals who are not currently receiving medications for these diseases.
Funding Source:
Government: Administration on Aging with cooperative agreement with Mecklenburg County Dept of Social Services
Reviewer Comments:

Limitations as cited by the authors:

  • Large set of exclusion criteria limits generalizability of the findings
  • Community-based sample resulted in very limited control over participant actions
  • Data collection ended after one year, which limited potential effects of lifestyle interventions that may be slow to produce impacts
  • Power analysis was not conducted on QALYs but clinical outcomes.
Quality Criteria Checklist: Review Articles
Relevance Questions
  1. Will the answer if true, have a direct bearing on the health of patients? Yes
  2. Is the outcome or topic something that patients/clients/population groups would care about? Yes
  3. Is the problem addressed in the review one that is relevant to dietetics practice? Yes
  4. Will the information, if true, require a change in practice? Yes
 
Validity Questions
  1. Was the question for the review clearly focused and appropriate? Yes
  2. Was the search strategy used to locate relevant studies comprehensive? Were the databases searched and the search termsused described? N/A
  3. Were explicit methods used to select studies to include in the review? Were inclusion/exclusion criteria specified andappropriate? Wereselectionmethods unbiased? N/A
  4. Was there an appraisal of the quality and validity of studies included in the review? Were appraisal methodsspecified,appropriate, andreproducible? N/A
  5. Were specific treatments/interventions/exposures described? Were treatments similar enough to be combined? Yes
  6. Was the outcome of interest clearly indicated? Were other potential harms and benefits considered? Yes
  7. Were processes for data abstraction, synthesis, and analysis described? Were they applied consistently acrossstudies and groups? Was thereappropriate use of qualitative and/or quantitative synthesis? Was variation in findings among studies analyzed? Were heterogeneity issued considered? If data from studies were aggregated for meta-analysis, was the procedure described? Yes
  8. Are the results clearly presented in narrative and/or quantitative terms? If summary statistics are used, are levels ofsignificance and/or confidence intervals included? Yes
  9. Are conclusions supported by results with biases and limitations taken into consideration? Are limitations ofthe review identified anddiscussed? Yes
  10. Was bias due to the review's funding or sponsorship unlikely? Yes