Testing normalization process theory in a randomized trial of mental health clinics implementing digital...

Normalization process theory (NPT) is one of the most highly cited implementation theories that explains the mechanisms by which new complex health interventions become embedded and sustained in healthcare settings; however, few of its predictions have been subjected to inferential hypothesis testing. In this theory-driven, ancillary analysis of a large hybrid type 3 effectiveness-implementation trial, we tested two NPT predictions: (1) its generative mechanisms of coherence, cognitive participation, collective action, and reflexive monitoring are modifiable in response to deliberate change efforts, and (2) greater enactment of these mechanisms predicts greater future sustainment of complex health interventions.

The trial tested two strategies to improve the implementation and sustainment of digital measurement-based care in outpatient mental health clinics serving youth. Twenty-one clinics were randomized to either training and technical assistance alone (k = 10) or training and technical assistance plus the Leadership and Organizational Change for Implementation (LOCI) strategy, in which leaders received training, coaching, and consultation to support implementation (k = 11). Six months after implementation strategies concluded, clinicians (N = 144) in both arms completed the Normalization MeAsure Development (NoMAD) questionnaire to describe the extent to which NPT mechanisms were enacted in their clinics. The primary outcome was a monthly, clinic-level, binary indicator of measurement-based care sustainment, derived from automatically-generated system usage data, for 16 months after the NoMAD assessment.

The NPT mechanisms were highly responsive to the organizational implementation strategy, which had a large effect overall (NoMAD total score: d_adj = 1.08, [0.63-1.52]) and on individual mechanisms of coherence (d_adj = 1.02, 95% CI = [0.60-1.44]), cognitive participation (d_adj = 1.00, [0.57-1.42]), collective action (d_adj = 0.96, [0.50-1.42]), and reflexive monitoring (d_adj = 1.01, [0.61-1.42]). Greater enactment of NPT mechanisms predicted significantly higher likelihood of measurement-based care sustainment in the month following the NoMAD assessment (adjusted event rate ratio for high versus low mechanism enactment = 2.96, [CI = 1.94-3.99]) and a significantly less steep decline in the log-odds of sustainment over the 16-month follow-up (b_adj = 0.32, SE = 0.15, p = .032).

The generative mechanisms proposed by NPT are modifiable in response to theoretically-aligned implementation strategies, and greater enactment of these mechanisms predicts greater sustainment of complex health interventions over 16 months.

ClinicalTrials.gov Identifier: NCT04096274 (Working to Implement and Sustain Digital Outcome Measures); Registered September 19, 2019; url: https://www.

gov/study/NCT04096274.