Author Affiliations: Division of Primary Care Internal Medicine, Departments of Medicine (Dr Takahashi) and Health Science Research (Drs Shah and Naessens), Mayo Clinic, Rochester, Minneapolis.
We appreciate the letters from other interested investigators. There are 3 major themes of discussion: patient selection, intervention, and baseline characteristics. Patient selection for telemonitoring remains a critical task for researchers or clinicians. We agree that patient selection encompasses an important decision when implementing telemonitoring. First, one chooses either to focus on a selected disease state like heart failure or to evaluate a risk-stratified population with multiple problems. We elected to focus on a set of patients with multiple comorbidities because we are clinically committed to the Minnesota medical home and are implementing practice changes for our high-risk populations. Berman et al discuss how one might stratify this population using risk predictors like nutritional status or functional status, which may better predict success for telemonitoring. Further research is needed for determining the best method for patient selection for medically complex patients in telemonitoring. Single disease state models like heart failure have the largest evidence base and benefit from well-prescribed treatment algorithms. The evidence for improvement of all-cause hospitalization for heart failure is mixed, with a modest improvement shown in a meta-analysis (risk ratio [RR], 0.91; 95% CI, 0.84-0.99),1 in which the largest randomized trial showed no difference in hospitalization.2
Takahashi PY, Shah ND, Naessens JM. Telehealth Monitoring With Nurse Clinician Oversight—Reply. Arch Intern Med. 2012;172(20):1612-1613. doi:10.1001/2013.jamainternmed.276