Telemedicine Evaluation

---
name: telemedicine-evaluation
description: Evaluating telehealth programs — study designs, quality metrics, equity, and implementation assessment.
category: scientific
---

## Overview

telemedicine-evaluation covers how to assess virtual care rigorously: whether telehealth delivers
equivalent outcomes, who it serves and who it excludes, and how to evaluate implementation.
It treats telehealth as a care-delivery intervention requiring the same evidentiary standards as
any other — not as a technology assumed beneficial.

## When to use

- Designing evaluations of telehealth programs (RCTs, stepped-wedge, quasi-experimental).
- Choosing outcomes: clinical equivalence, access, satisfaction, cost, equity.
- Non-inferiority designs for virtual vs in-person comparisons.
- Measuring the digital divide: who can't or won't use telehealth.
- Implementation evaluation: adoption, fidelity, workflow integration (RE-AIM, CFIR).
- Quality measurement: telehealth-adapted HEDIS/quality metrics.

## Core concepts

- **The right comparator.** Telehealth vs in-person for the same condition and population — not
  telehealth vs nothing. Non-inferiority designs are usually appropriate (margin: the largest
  clinically acceptable difference), since the claim is typically "as good, but more accessible."
- **Outcomes that matter.** Clinical outcomes (disease control, adverse events), access
  (wait times, travel burden, no-show rates), patient experience (validated satisfaction
  instruments, not smiley-face polls), cost (total cost of care, not just visit cost), and
  equity (uptake and outcomes by digital-access strata).
- **The digital divide as a validity issue.** Requiring video visits excludes patients without
  broadband, devices, or digital literacy — often the sickest and poorest. Measure reach by
  subgroup; audio-only options and digital navigators are equity interventions, not conveniences.
  An evaluation that only studies the already-connected overstates benefit.
- **Visit appropriateness.** Not everything suits virtual care: conditions needing physical exam,
  procedures, or delicate conversations need explicit triage criteria. Track conversion rates
  (virtual → in-person escalation) and missed findings as safety signals.
- **Implementation frameworks.** RE-AIM (Reach, Effectiveness, Adoption, Implementation,
  Maintenance) structures program evaluation; CFIR diagnoses implementation barriers. Use them —
  "we launched a platform" is not an evaluation.
- **Clinician experience.** Burnout, "webside manner" training needs, documentation burden, and
  reimbursement parity all determine sustainability. Measure clinician outcomes alongside
  patient ones.
- **Quasi-experimental options.** When randomization is infeasible: stepped-wedge rollouts,
  difference-in-differences around telehealth expansion, instrumental variables (distance to
  clinic). Each needs its assumptions stated and tested.
- **Sustainment.** Post-emergency evaluations often show regression when temporary
  reimbursement or enthusiasm fades. Evaluate at 12+ months with maintenance as an explicit
  RE-AIM dimension.

## Practical workflow

1. **Define the use case.** Which conditions, which visit types, which patients — with explicit
   virtual-appropriateness criteria.
2. **Choose the design.** RCT or stepped-wedge if possible; quasi-experimental with credible
   identification otherwise. Non-inferiority margins justified clinically.
3. **Measure comprehensively.** Clinical outcomes + access + experience + cost + equity strata.
   Prespecify the primary outcome.
4. **Track equity.** Uptake, completion, and outcomes by age, language, broadband access,
   socioeconomic status. Report who is missing.
5. **Assess implementation.** RE-AIM metrics; workflow integration; clinician burden;
   technical failure rates.
6. **Safety monitoring.** Escalation rates, missed diagnoses, emergency visits after virtual
   encounters — with chart review of signals.
7. **Report and sustain.** Full results including null/negative findings; maintenance plan;
   re-evaluation schedule as technology and reimbursement evolve.

## Common pitfalls

- Telehealth vs no-care comparisons inflating apparent benefit.
- Satisfaction surveys with no clinical outcomes ("patients liked it" ≠ "it worked").
- Ignoring the digitally excluded in both design and evaluation.
- No appropriateness criteria — everything virtual by default.
- Short evaluations during enthusiasm peaks (Hawthorne + novelty effects).
- Cost analyses counting only visit costs, missing downstream utilization.
- Clinician burnout and workflow disruption unmeasured.
- No-show rates and technical-failure rates unreported (hiding access barriers).
- Reimbursement assumptions baked into "cost-effective" claims without sensitivity analysis.
- Clinician training in virtual examination treated as optional rather than required.