Beyond Adoption: A New Framework for Theorizing and Evaluating Nonadoption, Abandonment, and Challenges to the Scale-Up, Spread, and Sustainability of Health and Care Technologies

This paper did not test a clinical intervention in the usual sense. Instead, it aimed to build and validate a framework — called **NASSS** (Nonadoption, Abandonment, Scale-up, Spread, and Sustainability) — to predict and explain why health and care technologies succeed or fail in the real world. The framework was developed by:

Synthesising 28 existing technology implementation frameworks from the research literature

Testing the framework against 6 real-world case studies of health technologies being introduced in the UK

The 6 technologies studied were:

Video outpatient consultations (Skype/FaceTime in hospital settings)

GPS tracking devices for people with cognitive impairment (e.g. dementia wandering)

Pendant alarm services (wearable emergency alert buttons)

Remote biomarker monitoring for heart failure (weight, blood pressure, oxygen saturation via tablet)

Care-organising software (web portals and apps for coordinating family caregiving)

Integrated data warehousing for case management across multiple NHS organisations

The comparator in each case was the status quo — existing care without the technology — or, for software products, different implementation approaches (e.g. co-designed app vs. solo-developer portal).

---

This is a mixed-methods systematic review combined with longitudinal qualitative case studies, not a clinical trial with a defined patient sample. However, the empirical fieldwork was substantial:

**Over 400 hours** of ethnographic observation across more than 20 organisations

**165 semi-structured interviews** with patients, clinicians, managers, technology developers, commissioners, and carers

**200 documents** (business plans, protocols, correspondence, clinical records)

Fieldwork conducted across **6 UK sites** over periods of up to **3 years** (2013–2017)

Settings ranged from large acute NHS hospital trusts to inner-London social care organisations, community hospitals, and national charities

Patient participants included people with heart failure, diabetes, cancer, cognitive impairment, and general frailty — predominantly older adults with multiple conditions and complex social circumstances

---

Because this is qualitative and framework-building research, there are no standardised psychometric scales or quantitative outcome measures. Instead, the researchers used:

**Ethnographic observation**: researchers embedded themselves in clinical settings, team meetings, home visits, and technology design sessions, recording field notes

**Semi-structured interviews**: open-ended conversations with all stakeholder types, audio-recorded and transcribed

**Document analysis**: reviewing business plans, protocols, ethics paperwork, and national policy documents

**"Go-along" interviews**: shadowing GPS device users with cognitive impairment in their daily routines

**Video recording**: both ends of video consultations captured for analysis

**Hermeneutic (interpretive) literature review**: not a standard PRISMA-style review, but a theoretically guided, iterative reading of ~28 technology implementation frameworks and thousands of citations

The final output was a 7-domain framework with 13 diagnostic questions, each answer classifiable as:

**Simple** — straightforward, predictable, few components

**Complicated** — multiple interacting components but ultimately manageable

**Complex** — dynamic, unpredictable, not decomposable into parts

---

**Study design:** Parallel mixed-methods — hermeneutic systematic review + longitudinal qualitative case studies (ethnography and action research), followed by iterative framework development and peer review.

**Why this design?** The authors explicitly argue that randomised controlled trials are the wrong tool for studying *why* technologies succeed or fail at scale. An RCT can tell you whether a technology works under controlled conditions; it cannot tell you why it is abandoned by 97% of patients outside those conditions, or why a hospital refuses to fund it after the trial ends. Qualitative longitudinal work can track the *process* of implementation — the workarounds, the political battles, the moments of collapse.

**Randomisation and blinding:** Not applicable. This is not a trial. One embedded case study (Case D, SUPPORT-HF) involved a randomised trial, but the NASSS paper analysed it ethnographically, not statistically.

**Duration:** Up to 3 years of fieldwork per case study, with some sites observed from 2013 to 2017. This long duration is a genuine strength — short evaluations routinely miss the abandonment phase.

**Literature search approach:** Non-standard. Rather than a database search with PRISMA protocol, the authors used "ancestry and snowballing" — tracking citations from key papers and manually screening ~7,500+ titles. They acknowledge this is less reproducible but argue that standard database searches were "neither sensitive nor specific" for this literature.

**Framework refinement:** The draft NASSS framework was shared with colleagues running 10 additional technology programmes (including primary care email consultations, mental health peer support networks, and remote transplant monitoring) and refined based on feedback — a form of external validity testing, though not formal empirical validation.

**What this design can prove:**

That the 7 NASSS domains are theoretically grounded and empirically recognisable across diverse technology programmes

That complexity in multiple domains simultaneously is associated with failure to mainstream

**What this design cannot prove:**

Causal relationships — it cannot prove that fixing domain X will cause technology Y to succeed

Generalisability beyond UK health and social care settings (though the authors argue the framework is broadly applicable)

Statistical effect sizes — there are no quantitative predictions, e.g. "technologies with 4+ complex domains have a 70% abandonment rate"

Predictive validity — the framework had not been prospectively validated in an independent sample at time of publication

**Major methodological weaknesses:**

The hermeneutic literature review is not reproducible in the way a PRISMA review is; other researchers might identify different foundational frameworks

The case studies were not selected randomly — they were drawn from two pre-existing research programmes the authors were already running, introducing selection bias

The framework was refined partly by the same team that developed it, raising risk of confirmation bias

No quantitative outcome data, so the claim that "complexity in multiple NASSS domains rarely leads to mainstreaming" is an interpretive judgment, not a statistic

---

**From the literature review:**

28 technology implementation frameworks were identified; 14 took a dynamic systems approach

No prior framework focused explicitly on nonadoption, abandonment, or sustainability — most studied short-term adoption only

Most frameworks ignored: the heterogeneity of the patient's condition; the symbolic meaning of technologies (e.g. GPS trackers connoting loss of autonomy); health inequalities in access; the organisation's business model; and long-term resilience of the implementation

**From the NASSS framework itself — 7 domains:**

**Domain 1 — The Condition:** Is the illness simple, complicated, or complex in its clinical presentation, comorbidities, and sociocultural dimensions? In most of the studied cases, only a minority of patients were deemed clinically suitable for the technology

**Domain 2 — The Technology:** Four sub-questions covering (2A) material/technical features and dependability; (2B) what knowledge the technology generates and whether users trust it; (2C) what knowledge and support users need to operate it; (2D) the supply model and risk of vendor lock-in or market withdrawal

**Domain 3 — The Value Proposition:** Does the technology create value for patients, clinicians, managers, investors, and regulators — and are these value perceptions aligned? In practice, upstream investor value and downstream patient/clinician value were frequently misaligned

**Domain 4 — The Adopter System:** Three sub-domains — professional staff (including clinician resistance rooted in professional identity, not just usability); patients (including health literacy, digital literacy, burden of self-management); and lay caregivers (informal support networks)

**Domain 5 — The Organisation(s):** Business model, financial sustainability, fit with existing routines, staff capacity for implementation work, and organisational resilience

**Domain 6 — The Wider Context:** Policy, regulation, interoperability standards, legal liability, market dynamics, and societal factors

**Domain 7 — Interaction and Adaptation Over Time:** Whether the programme can adapt as the environment changes; whether complexity compounds across domains

**From the empirical case studies:**

Video consultations: