Does investment in a telephonic diabetes intervention result in significant decreases in hospitalizations and costs?
As payment for healthcare services transitions from a fee-for-service model to a value-based model as seen with Medicare, Medicaid, and other third-party payers that have begun to incorporate this into their services, the methods by which the quality of care is measured must also transition. In a fee-for-service model, a patient with type 2 diabetes (T2D), for instance, would have a visit with their provider to receive care, and the provider would get reimbursed for the care. In the same situation in a value-based model, the reimbursement to the provider would be contingent on quality measurements, such as the patient’s HBA1c. Recent work has been done to develop and assess new measurements of quality care and also to incorporate innovative interventions to determine their capacity to improve both quality and cost-effectiveness of care.
One such study with this aim was the Bronx A1C study, which was a randomized controlled trial comparing a diabetes self-management intervention of print-only instructions versus telephonic behavioral interventions in addition to the printed materials over one year. The primary outcome of the Bronx A1C trial was HBA1c change. In the telephonic intervention arm, a significant decrease was seen in HBA1c in the patient population with baseline HBA1c >9%, but a similar result was not shown in patients with baseline HBA1c between 7.1% and 9.0%. In addition to lowering HBA1c, other outcomes are of interest to be incorporated into value-based healthcare metrics.
This study, by Tabaei et al., examined the long-term effect of the Bronx A1C trials one-year telephonic diabetes self-management intervention on health care utilization and cost-effectiveness. The researchers used statewide health databases to collect data on all-cause hospitalizations, diabetes-related hospitalizations, and ED services for the participants of the Bronx A1C trial. The study period was up to two years before enrollment / up to four years post-enrollment. The collection of data before enrollment allowed the researchers to examine preintervention rates of the outcomes. Diabetes-related hospitalizations were categorized into short-term complications, long-term complications, and lower-extremity amputations, each considered to be related to diabetes if the diagnosis also included diabetes. Data were analyzed using two-tailed t-tests and Wilcoxon rank-sum tests for continuous variables and chi-squared tests for categorical variables.
Preintervention rates of inpatient hospitalization, diabetes-related hospitalization, and ED use were similar between the telephonic intervention and print-only groups. During the four-year study follow-up, all-cause hospitalizations were significantly different between the groups with the telephonic intervention group showing decreased all-cause hospital utilization (OR 0.89, CI: 082-0.97, p<0.01). An even more substantial effect was seen in the diabetes-related hospitalizations in the telephonic intervention group compared to the print-only group in the four-year follow-up (OR 0.83, CI: 075-0.93, p<0.001). Costs followed the same trends as hospitalization rates, but an increase in outpatient and pharmacy use was observed in both the groups that received any self-management intervention. Interestingly, when the researchers compared the HBA1c reduction between the groups in the Bronx A1C trial to hospital utilization, they found no significant correlation. Also of interest, when comparing the group with a baseline HBA1c of 7% to 9% in comparison to the >9% group, they found that the >9% group who received the telephonic intervention had a significant reduction in their diabetes-related hospitalization rates, but not their all-cause hospitalization rates. In contrast, the 7%-9% group had significant reductions in both. Concerning monetary savings, it was found that 33 patients needed to receive the telephonic intervention to prevent one hospitalization in comparison to the print-only group. The cost to provide telephonic self-management intervention to 33 patients was calculated to be $8,903, whereas the cost of hospitalization was $10,703, creating a cost-benefit of $1,800.
The results showed that any intervention, whether telephonic or print-only, was better at decreasing hospital utilization than no intervention at all. Comparing hospital utilization in the telephonic intervention group to the print-only group yielded no surprises, showing significant reductions in the telephonic intervention group. While these results are intuitive, less intuitive is the result showing that HBA1c reduction was not significantly correlated to decreased hospital utilization. In other words, how much an intervention lowers HBA1c does not necessarily equate to decreases in hospitalization and cost benefits. Also of interest, the study did show a significant decrease in diabetes-related hospitalization for the >9% HBA1c group between the two study arms, but not in all-cause hospitalization. There are many possible explanations for this, such as those with higher baseline HBA1c being more likely to have comorbidities or being older, thus leading to increases in non-diabetic related hospitalizations. Still, it highlights the importance of the multicomponent approach to diabetes in that HBA1c should not be the only focus. Overall, the study provided evidence that a telephonic intervention decreased costs and rates of hospital utilization in diabetes patients.
- Telephonic diabetes self-management interventions significantly decreased hospital utilization and costs in comparison to only printed materials
- HBA1c reduction did correlate to a decrease in hospital utilization over the four-year follow-up
- Any self-management intervention, whether telephonic or print-only, yielded significant improvements over no intervention
Tabaei, Bahman P, et al. “Impact of a Telephonic Intervention to Improve Diabetes Control on Health Care Utilization and Cost for Adults in South Bronx, New York.” Diabetes Care vol. 43,4 (2020): 743-750. doi:10.2337/dc19-0954
David Clarke, PharmD Candidate, University of Colorado, Skaggs School of Pharmacy and Pharmaceutical Sciences