Implementation science (IS) is the study of why things that work in trials often don't work in the real world, and what to do about it. It is not asking "does this intervention work?" That is efficacy research, and we already have the answers for most cancer control interventions. IS asks "why isn't it working here, for these people, in this health system?"
Peters et al. (2013) describe implementation research as investigating the "what, who, how, where, and why of putting evidence into practice." The WHO Alliance for Health Policy and Systems Research (2013) frames it as research that generates knowledge to translate findings into effective, scalable health programmes, language that resonates directly with African Ministry of Health decision-makers.
For cancer control in Africa, the evidence base is solid. HPV vaccines, HBV vaccination, VIA-based cervical screening, and FOLFOX chemotherapy are all proven. The health system capacity to actually deliver them to everyone who needs them — that is where things fall apart.
The CFIR is the most widely used IS framework in the world. Think of it as a structured way to ask: where exactly is the problem? The original version (Damschroder et al., 2009) organised barriers and facilitators across five domains. In 2022, Damschroder and colleagues published CFIR 2.0 — a significant update that sharpens the tool and, for the first time, bakes equity explicitly into the analysis. That matters a great deal in Africa, where the people most in need of cancer interventions are routinely the last to receive them.
The key update: Intervention Characteristics is now called Innovation, signalling that CFIR applies to any change in practice, not just a clinical protocol. The framework now asks not only who receives an intervention, but who doesn't — and why.
| CFIR 2.0 Domain | What it examines | African Cancer Example |
|---|---|---|
| Innovation (formerly Intervention Characteristics) | Evidence strength, adaptability, trialability, complexity, equity of the innovation | Single-dose HPV schedule perceived as unproven; cryotherapy equipment reliability |
| Outer Setting | National policy, financing, partnerships, equity considerations, community needs | No funded national cervical cancer strategy; GAVI transition risk; Africa CDC cancer plan |
| Inner Setting | Facility culture, infrastructure, leadership, readiness for change | Non-functional colposcope; no maintenance budget; cancer not a clinical priority |
| Individuals | Knowledge, beliefs, self-efficacy, personal attributes of providers | Nurses trained in VIA years ago — no refresher, low confidence |
| Implementation Process | How the programme is planned, launched, run, and improved | HPV programme launched without community sensitisation; no adaptation when uptake was low |
The key rule: CFIR 2.0 is a diagnostic tool, not a checklist. Identify the dominant barriers first — then match your strategies to them. Skipping the diagnosis is why most implementation efforts fail.
Here is a mistake that derails many cancer programmes: confusing clinical outcomes (did cervical cancer incidence fall?) with implementation outcomes (is the programme actually being delivered?). You cannot fix the first without measuring the second. Proctor et al. (2011) gave the field eight precise implementation outcomes, each describing a different way a programme can succeed or fail:
| Outcome | Plain definition | Cancer example |
|---|---|---|
| Acceptability | Do people find the intervention agreeable? | Do women in this community actually want VIA screening? |
| Adoption | Did facilities or providers decide to use it? | Has the district hospital started offering LEEP? |
| Appropriateness | Does it fit this setting and context? | Is FIT realistic in this primary care facility? |
| Feasibility | Can it actually be carried out here? | Do we have the equipment and trained staff for cryotherapy? |
| Fidelity | Is it being delivered as intended? | Are screen-positive women treated at the same visit, or sent away? |
| Implementation Cost | What does delivery actually cost? | Cost per woman screened when VIA is integrated into ANC |
| Penetration | Is it reaching the whole system, reaching across the whole system? | What % of eligible women across all district facilities are being screened? |
| Sustainability | Is it still running after the pilot ends? | Do HPV vaccination rates hold up after GAVI support transitions to the national budget? |
Why does this matter? If a cervical screening programme fails, you need to know why. A fidelity failure (women are screened but not treated at the same visit) demands a completely different response than an acceptability failure (women are choosing not to attend). Without this taxonomy, programmes repeat the same intervention and expect different results.
By the end of this course you will know CFIR, COM-B, ERIC, RE-AIM, and the Dynamic Sustainability Framework. The most common question people ask after that is: which one do I actually reach for in practice? Here is the answer as a single table.
| The question you are asking | The framework to use |
|---|---|
| What is blocking implementation — where exactly is the problem? | CFIR 2.0 — diagnose barriers across Innovation, Outer Setting, Inner Setting, Individuals, and Process |
| Why are providers or communities not changing their behaviour? | COM-B — identify whether the gap is Capability, Opportunity, or Motivation |
| What specific strategies should we use to address those barriers? | ERIC — select from 73 implementation strategies grouped into 9 clusters, matched to your diagnosed barriers |
| Did implementation actually work, and for whom? | RE-AIM — evaluate Reach, Effectiveness, Adoption, Implementation fidelity, and Maintenance |
| How do we keep this running after the pilot ends? | Dynamic Sustainability Framework — plan sustainability into the programme design from day one, not after the donor leaves |
These frameworks are not competing — they are sequential. CFIR tells you where to look. COM-B tells you what kind of problem it is. ERIC tells you what to do about it. RE-AIM tells you whether it worked. And the Dynamic Sustainability Framework stops you from having to start again in five years.
This course works primarily with CFIR 2.0, ERIC, COM-B, and RE-AIM. But three other frameworks come up regularly in African cancer implementation work and are worth knowing:
| Framework | What it does | When to use it in cancer control | Reference |
|---|---|---|---|
| Normalisation Process Theory (NPT) | Explains why new practices do — or don't — become embedded in everyday routine | When a VIA programme is technically running but slowly dying because it never became part of how the facility actually works. NPT's four constructs — coherence, cognitive participation, collective action, reflexive monitoring — pinpoint where the embedding broke down. | May C et al. Implement Sci. 2009;4:29. |
| Theoretical Domains Framework (TDF) | Identifies the behavioural root cause of a health worker practice problem | When nurses aren't performing VIA or administering the HepB birth dose. You need to know whether the problem is knowledge, skills, beliefs, motivation, or social norms. Use it alongside COM-B. | Cane J et al. Implement Sci. 2012;7:37. |
| Health Equity Implementation Framework | Adds an equity lens to every CFIR domain | When you need to ask not just "why isn't the programme working?" but "who specifically is it not reaching, and why?" Developed for marginalised populations — fits the African cancer context directly. | Woodward EN et al. Implement Sci. 2021;16:13. |
The science here is unambiguous. The PATRICIA trial showed 92.9% efficacy against CIN2+ for HPV 16/18 in HPV-naive women (Paavonen et al., Lancet 2009). More recently, the KEN SHE trial (Barnabas et al., NEJM Evidence 2022) confirmed that a single dose is non-inferior to two doses in girls aged 9–14. WHO updated its position accordingly in 2022. That is a significant programmatic win. Eliminating the recall visit removes one of the biggest barriers to completion in resource-limited settings.
And yet HPV vaccine coverage in sub-Saharan Africa sits below 20%, against WHO's 90-70-90 target of 90% by 2030. This is not a vaccine problem. It is a delivery problem.
The girls with the lowest coverage: out-of-school girls, those in rural areas, and those from the poorest households. They are also the ones with the highest cervical cancer risk. CFIR 2.0 forces us to name this as an equity failure, not just a logistical shortfall. ERIC responses include: community champion mobilisation, religious leader engagement, community-based delivery beyond schools, and multi-sector coordination linking health, education, and community systems.
A common failure pattern in African cancer prevention programmes: design the intervention, train the health workers, launch the programme, and then wonder why communities aren't showing up. Co-design flips this. It means involving community members, leaders, and potential users in shaping the programme before it launches, not after uptake has stalled.
Rwanda's HPV vaccination programme reached over 90% coverage within three years. A key enabler was working with community health worker networks and women's groups from the start, not parachuting in a programme and expecting it to run. Uganda's experience with cervical cancer screening shows that involving traditional and religious leaders as implementation champions (a CFIR 2.0 Inner Setting move) substantially increases uptake in communities where those leaders shape health behaviour.
CFIR 2.0's Implementation Process domain treats engaging, meaning formally involving affected communities in programme planning, as a required step, not optional outreach. The evidence supports it: participatory-designed programmes in sub-Saharan Africa consistently outperform top-down ones on sustainability.
Here is a tension that comes up in every cancer programme that moves from a research setting into routine care: the original protocol was designed in one context, and you are delivering it in a different one. Do you stick rigidly to the original design, or do you adapt it to fit your setting? The answer is neither extreme.
Core components are the active ingredients of an intervention, which are the elements that actually drive the outcomes. Changing them undermines effectiveness. Adaptable periphery refers to everything else: delivery format, language, timing, the cadre of staff delivering it, the platform it sits on. Adapting the periphery improves fit without sacrificing what works. The practical skill is knowing which is which. That requires understanding the theory behind the intervention, not just the protocol.
The FRAME (Framework for Reporting Adaptations and Modifications to Evidence-Based Interventions, Stirman et al., 2019) gives practitioners and researchers a structured way to document what was changed, why, who decided, and whether the change was consistent with the core intent of the intervention. This matters in African cancer programmes for a specific reason: adaptations happen whether you document them or not. CHWs modify delivery scripts, nurses adjust screening thresholds to match available equipment, health educators translate materials informally. Without systematic documentation using something like FRAME, those adaptations are invisible. You cannot learn from them, replicate the good ones, or correct the harmful ones.
In practice: adapt the platform (integrate cervical screening into an existing ART visit rather than a separate visit). Protect the core (maintain same-visit treatment. Do not remove the treatment step to make delivery easier). Document every change. Evaluate the outcome.
About 85–90% of hepatocellular carcinoma in sub-Saharan Africa is caused by chronic HBV infection acquired at or around birth (Llovet et al., 2021). The HBV vaccine has been available since 1986 and provides ≥95% seroprotection. A 20-year cohort study in Taiwan: a high-HBV setting comparable to West Africa. It showed an 84% reduction in HCC incidence in vaccinated cohorts (Chang et al., NEJM 1997). That is compelling evidence. Note that Taiwan is a middle-income country; the biological proof of concept is robust, even if health system comparisons require care.
The critical intervention is the hepatitis B birth dose (HepB-BD), given within 24 hours of delivery. Without it, a complete three-dose series starting at 6 weeks is too late to interrupt perinatal infection. Most perinatally infected infants develop chronic HBV and carry a lifetime HCC risk that plays out 30–50 years later. Despite this, birth-dose coverage across much of sub-Saharan Africa remains below 20%.
The barriers are concrete. Most births happen at home (Outer Setting — access). Facilities lack integrated postpartum-immunisation protocols (Inner Setting). Health workers are not trained in newborn vaccine administration (Individuals). There is no community pathway for home births (Process). Each barrier has a matched ERIC response. None of them require a new vaccine. They require fixing the delivery system.
There is no vaccine for colorectal cancer. Prevention means changing what people eat and how they live. That is a harder implementation challenge than vaccines. The evidence on risk factors is solid: 10g/day more dietary fibre lowers CRC risk by about 10–20%; each 50g/day increment of processed meat increases risk by 18%; physical activity cuts risk by roughly 20–25% for the most active versus least active individuals (World Cancer Research Fund/AICR, 2018).
On aspirin: it was once considered a viable chemoprevention option. Randomised trial data (Chan et al., NEJM 2012) showed 20–40% reduction in adenoma recurrence. But the ASPREE trial (McNeil et al., NEJM 2018) — 19,114 adults aged ≥70, randomised to 100mg aspirin or placebo — found aspirin was associated with a higher rate of cancer mortality (HR 1.31; 95% CI 1.10–1.56). That finding substantially changes the risk-benefit calculation, particularly for older adults. Aspirin is no longer recommended as a population-level prevention strategy in African contexts without careful individual clinical assessment.
For behavioural prevention, COM-B is the right tool. The question is not just whether people know about CRC risk — it is whether they have the capability (knowledge), opportunity (affordable fibre-rich food, safe places to exercise), and motivation (CRC feels like a real threat) to change. Awareness campaigns address only one of three levers.
WHO's 2021 guidelines put HPV DNA testing first (sensitivity 87–96% for CIN2+), with VIA as the alternative where HPV testing infrastructure doesn't exist. Both are valid. But the choice of test is almost secondary to this single implementation principle: treat at the same visit.
The data are consistent across studies in sub-Saharan Africa: when treatment requires a separate visit, 30–60% of screen-positive women never come back. They don't decline treatment. They run out of transport money, or they have children to care for, or the referral hospital is two hours away, or the appointment system doesn't work. Screen-and-treat closes this gap by training the same provider who does the screening to also perform cryotherapy or thermocoagulation for eligible lesions on the spot. It is not a compromise. It is the evidence-based approach for this context, and it is what WHO recommends (2021).
If you work in East or Southern Africa, this section is essential. Women living with HIV (WLHIV) have a 5-fold elevated risk of cervical cancer compared to HIV-negative women. HIV accelerates the progression from HPV infection to invasive cancer, meaning cancer appears years earlier (de Vuyst et al., Int J Cancer 2008). In East Africa, cervical cancer in HIV-positive cohorts regularly presents in women in their late 20s and early 30s. In Southern Africa, where HIV prevalence among women of reproductive age exceeds 30% in some settings, HIV is the dominant driver of the cervical cancer burden.
The standard recommendation to begin screening at age 30 does not apply to WLHIV. WHO 2021 recommends starting at age 25 (or at first HIV diagnosis, whichever comes first), and repeating every 3–5 years rather than every 5–10 years. The most practical implementation approach: integrate cervical screening into ART clinic visits. She is already there, already engaged in care, and the visit adds minimal burden.
West and Central Africa face lower HIV-related cervical cancer burden, but different outer-setting challenges: fragmented health financing, OHADA legal frameworks governing health facilities in francophone countries, and limited cross-sector coordination capacity.
Most cancer implementation analyses ask: why isn't the programme working? Equity implementation asks a harder question: for whom isn't it working, and what structural reasons explain that? These are different questions with different answers and different strategies.
The Health Equity Implementation Framework (HEIF) (Woodward EN et al., 2021) extends CFIR by adding equity as an explicit analytical layer at every domain. Where CFIR asks "what barriers exist in the Outer Setting?", HEIF asks "which barriers fall disproportionately on specific groups, specifically asking which groups are defined by race, gender, poverty, geographic isolation, or HIV status?" In the African cancer context, this is not an abstraction. Out-of-school girls face higher cervical cancer risk and lower HPV vaccine coverage than their peers. Women in rural areas face longer distances to any care. Women living with HIV face faster disease progression. Each of these is a structural determinant, not an individual preference.
Intersectionality refers to the overlapping and compounding effects of multiple disadvantages. This matters in practice. A rural, poor, HIV-positive woman in Malawi faces a different implementation barrier profile than an urban, employed woman in Nairobi. Treating "sub-Saharan Africa" as a single implementation context, or reporting aggregate coverage without equity stratification, obscures who is being left behind. CFIR 2.0's equity constructs and the HEIF framework both prompt programme managers to stratify data by wealth quintile, urban/rural location, HIV status, and education from the start, not as an afterthought when aggregate targets are met.
Structural determinants — the social, economic, and political conditions that systematically disadvantage certain groups cannot be fixed by a better demand-generation campaign. They require ERIC strategies in the "Change Infrastructure" and "Develop Stakeholder Interrelationships" clusters: policy advocacy for financial protection, community-based delivery that reaches the hardest-to-reach, and integration with social protection programmes.
The faecal immunochemical test (FIT) is well-suited for African primary care: no bowel prep, no dietary restriction, no clinical contact for sample collection, pooled sensitivity of approximately 79% for CRC with 94% specificity, though these figures vary by threshold and manufacturer (Lee et al., Ann Intern Med 2014). The implementation problem is not the test. It is that most African countries have no national CRC screening programme at all. Before designing a clinical programme, programme managers need to address the Outer Setting (CFIR 2.0), building the policy, the guidelines, and the financing commitment that make a programme possible. That is the real first step.
For chronic HBV carriers, abdominal ultrasound every 6 months with or without AFP reduces HCC mortality by 37–45% by catching tumours at resectable stages (Singal et al., Ann Intern Med 2014). On AFP thresholds: the AASLD/EASL standard is ≥20 ng/mL. In settings where ultrasound is unavailable, some programmes use a ≥200 ng/mL threshold, accepting lower sensitivity to avoid swamping limited referral capacity with false positives. This is a legitimate, documented adaptation — not a gap. Name it as such in your programme design.
A practical note for programme managers: most DHIS2 implementations in Africa do not have cancer-specific indicators. HBsAg testing rates, surveillance uptake, and HCC detection rates are rarely captured in routine data systems. If you want to know whether your programme is working, build the data indicators into your DHIS2 module from day one, not as an afterthought when an evaluation is commissioned.
Between 60–80% of African cancer patients arrive with advanced-stage disease (Pace & Shulman, J Clin Oncol 2016; Stefan, Ecancermedicalscience 2015). This is a system failure, not a personal failure of patients. COM-B (Michie et al., 2011) helps diagnose exactly what is going wrong and why. The model identifies three levers: Capability, Opportunity, Motivation. Late presentation in African cancer care involves all three, and most programmes address only one:
| Lever | What is failing | What actually helps |
|---|---|---|
| Capability | People don't recognise cancer symptoms or don't know treatment is possible | CHW symptom screening; cancer literacy in primary care |
| Opportunity | Can't afford transport; no nearby facility; financial catastrophe from treatment costs | Transport vouchers; decentralised services; patient navigation; financial protection |
| Motivation | Fatalism — "cancer means death"; fear; stigma; preference for traditional medicine | Survivor champion networks; positive framing; religious leader engagement |
A transport voucher does nothing for someone who believes cancer is untreatable. A survivor champion does nothing for someone who cannot afford the bus. Diagnose the barrier type first, then match the strategy.
If there is one implementation strategy that African health systems have returned to more than any other, it is this one. Task-shifting, moving specific clinical tasks from higher-cadre to lower-cadre health workers, and this is how most African cancer programmes actually function at scale, even if they do not always name it as such.
The evidence base is solid. A systematic review of task-shifting in cancer control in LMICs found that tasks including screening, patient education, and navigation were performed comparably, and in some cases better, than by non-physician health workers than by physicians (Butler et al., J Glob Oncol 2023). For cervical cancer specifically, nurse-led VIA and same-visit cryotherapy is now WHO-recommended practice. CHW-delivered navigation in South Africa reduced loss to follow-up from 21% to 6% at six-month visits (Adewole et al., 2017). In Nigeria, community health extension workers (CHEWs) trained in VIA achieved knowledge scores rising from 52% to 92% post-training and screened effectively in rural communities.
Three task-shifting patterns that matter most in African cancer implementation:
- Nurse-led screening and treatment: Nurses trained in VIA and cryotherapy or thermocoagulation remove the physician bottleneck from the cervical screening pathway entirely. This is the core of WHO's screen-and-treat recommendation. It is not just a clinical preference, it is an implementation strategy for a health system with too few doctors.
- CHW patient navigation: Community health workers who accompany patients through the care pathway, helping them navigate appointments, transport, cost barriers, and anxiety, substantially reducing loss to follow-up between diagnosis and treatment. Navigation is an ERIC "Develop Stakeholder Interrelationships" strategy that works at the individual level.
- Decentralised treatment follow-up: Moving post-treatment monitoring and supportive care from tertiary hospitals to district facilities or primary health centres. With teleconsultation support from specialists, this approach allows more patients to complete treatment without repeated long-distance travel. James, the colorectal cancer patient in the Navigator game, needed exactly this.
The CFIR 2.0 lens: task-shifting addresses Outer Setting barriers (workforce shortage, geographic access) and Inner Setting barriers (no physician available) simultaneously. It requires ERIC strategies in "Train and Educate Stakeholders" and "Change Infrastructure." The risks: scope creep, inadequate supervision, and quality erosion. These are real and require built-in competency assessment and structured mentoring, not just a one-time training.
Cervical cancer: This course uses FIGO 2018 staging, which added imaging and pathological criteria to the older 2009 system. Stage IIB (parametrial involvement) is not surgically curable — concurrent cisplatin-based chemoradiotherapy is the standard (Green et al., Cochrane 2005). For early-stage disease (IA–IIA), open radical hysterectomy is now the preferred surgical route following the LACC Trial (Ramirez et al., NEJM 2018), which found minimally invasive surgery was associated with significantly worse disease-free survival (86% vs 96.5% at 4.5 years). Fewer than 30 of 54 African countries have any radiotherapy services. That is the implementation gap.
Colorectal cancer: Adjuvant FOLFOX remains standard for stage III. But the IDEA collaboration (Grothey et al., NEJM 2018), pooling six randomised trials, showed that 3 months of CAPOX is non-inferior to 6 months for low-risk stage III CRC. In African settings where chemotherapy completion rates are poor, this matters: three-month regimens are far more likely to be completed. They should be the default where feasible.
HCC: Atezolizumab + bevacizumab (IMbrave150, Finn et al., NEJM 2020) and the HIMALAYA trial's durvalumab + tremelimumab (STRIDE regimen) are the current first-line standards for advanced HCC. Sorafenib remains an option (SHARP trial, Llovet et al., NEJM 2008). But originator sorafenib costs $5,000–9,000 per month, effectively inaccessible in most African public health systems. Generic sorafenib exists but supply is unreliable. The honest implementation answer for HCC in Africa is still this: prevent it with HepB vaccination. Once advanced HCC is diagnosed, treatment access in most African settings is extremely limited.
Because most African cancer patients present at stage III–IV, palliative care is the most commonly needed cancer intervention on the continent. It is also the most consistently under-resourced. The WHO designates palliative care as a core component of universal health coverage, not an optional specialty. The African Palliative Care Association (APCA) has developed Africa-specific standards and implementation frameworks that are freely available and contextually grounded.
Four implementation priorities that make the biggest difference:
- Oral morphine access: WHO lists oral morphine as an essential medicine. In most of sub-Saharan Africa, it is still nearly impossible to access, blocked by regulatory barriers, prescribing restrictions, and procurement failures. ERIC "Change Infrastructure" means advocating directly for opioid policy reform and fixing the supply chain.
- Task-shifting to primary care: Uganda, Kenya, and Zimbabwe have demonstrated that trained nurses and CHWs can deliver basic palliative care effectively. Uganda is a globally recognised model for integrating community-based palliative care into primary health.
- Early integration with treatment: Temel et al. (NEJM 2010) showed that early palliative care integration alongside oncological treatment improves quality of life and, in some contexts, survival. "Palliative" does not mean "giving up" — it means treating suffering alongside disease.
- CFIR 2.0 barriers: Outer Setting (opioids excluded from national insurance; no palliative care policy); Inner Setting (no protocols, no team at district level); Individuals (fear of opioid prescribing among health workers).
RE-AIM (Glasgow et al., 1999) is the evaluation framework that prevents the most common self-deception in cancer control: reporting good clinical outcomes from a small, well-funded pilot and calling it a success. A programme with 90% efficacy but 15% reach delivers 13.5% of its potential population benefit. RE-AIM makes that visible.
| Dimension | The question | The equity question | Cancer example |
|---|---|---|---|
| Reach | What proportion of the target population actually received it? | Are the most marginalised groups (rural, poor, out-of-school, WLHIV) reaching the same rates? | HPV coverage broken down by wealth quintile and school enrolment, not just aggregate figures |
| Effectiveness | What is the real-world impact in this setting? | Are outcomes equitable across subgroups? | VIA sensitivity under operational (not trial) conditions; screen-and-treat completion rates |
| Adoption | What proportion of eligible settings and providers are using it? | Are lower-resource facilities adopting at the same rate? | What % of district hospitals offer LEEP? Are all trained nurses performing VIA in practice? |
| Implementation | Is it being delivered as intended? What adaptations were made? | Are equity-targeted adaptations documented and evaluated? | Is screen-and-treat actually happening at the same visit, or deferred? |
| Maintenance | Is it still working six to twelve months later? Is it institutionalised? | Do rural programmes lose funding before urban ones? | HepB birth-dose rates after GAVI transition; HPV coverage after the external project ends |
Source: Glasgow RE et al. Am J Prev Med. 1999;17(1):51–56. | Kessler RS et al. Implement Sci. 2013;8:110.
One of the biggest inefficiencies in global health research is running a clinical effectiveness trial, then years later running a separate implementation study to figure out how to deliver what was just proven to work. Hybrid effectiveness-implementation trials collapse this pipeline. Curran et al. (Med Care, 2012) defined three types:
| Type | What it tests | Cancer example |
|---|---|---|
| Type 1 | Clinical effectiveness is the primary question; implementation data collected alongside | RCT of HPV vaccine schedules, with concurrent data on delivery barriers and provider behaviour |
| Type 2 | Both clinical and implementation strategies are tested simultaneously | Testing screen-and-treat protocol effectiveness AND the nurse training package needed to deliver it at the same time |
| Type 3 | Implementation strategy is the primary question; clinical outcomes monitored | Testing a CHW-based HBsAg testing campaign while tracking HCC detection rates as the downstream outcome |
It is also worth separating implementation research from quality improvement (QI). QI uses rapid iterative cycles to improve a specific local process: valuable for local improvement, but not designed to generate knowledge applicable elsewhere. Implementation research produces generalisable findings. African cancer programmes need both, at different times and for different questions.
The pilot-to-scale gap in African cancer control is not a mystery. It is predictable. Pilots concentrate resources, attention, and motivated staff in one place. National roll-out distributes all of those across a much larger, more heterogeneous system. The things that made the pilot work: close supervision, enthusiastic champions, reliable supply chains, and fast problem-solving — are often not built into the national plan.
Milat et al. (2013) identify the recurring failure points: fidelity erodes as ownership shifts from researchers to routine health workers; sustainability is not planned until external funding is about to end (too late); adaptations that are actually necessary are resisted as protocol deviations; and the outer setting, including policy, financing, and workforce systems, which were never properly engaged at the outset.
The Dynamic Sustainability Framework (Chambers et al., 2013) is direct about this: sustainability is not what happens after implementation succeeds. It must be designed into implementation from day one. For African cancer programmes that means: build domestic financing transition timelines before the donor comes in; embed HPV vaccination in the national immunisation schedule, not just a project plan; include DHIS2 indicator development in the budget; and build community demand so the programme can survive staff turnover.
On the economics: the Disease Control Priorities 3rd edition (DCP3) estimates cervical cancer prevention (HPV vaccination + VIA screening) at $10–100 per DALY averted, among the most cost-effective cancer interventions in the world. HepB vaccination for HCC prevention is in the same range. When you present these numbers to a Ministry of Finance alongside the WHO NCD Global Monitoring Framework targets and the Africa CDC Africa Health Strategy, the investment case becomes very hard to argue against.
- Use CFIR 2.0 to diagnose where a cancer programme is failing — and why
- Match ERIC implementation strategies to specific barriers in HPV vaccination and HepB birth-dose delivery
- Apply the screen-and-treat principle and explain why the referral step kills people in the African context
- Use COM-B to identify the right type of strategy for a late-presentation problem: structural, motivational, or knowledge-based
- Evaluate a cancer programme using all five RE-AIM dimensions, including the equity sub-questions
Intervention · Outer Setting · Inner Setting · Individuals · Process
Engage · Train · Change Infrastructure · Evaluate · Develop Relationships
Reach · Effectiveness · Adoption · Implementation · Maintenance
Capability · Opportunity · Motivation → Behaviour