Narrative structure of primary health care investment actions

Structuring exercise — five domains · Working document — September 2026

How these texts are organised

Each text covers a single investment action and occupies one page. It is built from nine blocks. The first four form a continuous paragraph, with the markers shown in bold and in square brackets at the point where each function ends. The remaining blocks are set out separately.

1
Situation

facts the reader accepts without argument, placing author and reader at the same starting point.

2
Complication

what breaks that situation and forces a decision.

3
Reason for sequencing

why this action comes before or after the others.

4
Answer

the proposed investment action, in one sentence.

5
Possible actions

options between which the decision-maker chooses, differing in ambition and cost. These are orientations, not a recommendation.

6
Actions to avoid

plausible paths that do not lead to the intended result.

7
Synergy with the other levers

what this action unlocks in other levers, shown as a diagram.

8
Order of magnitude of cost

the nature of the cost and what determines its size.

9
What we expect to see

the horizon at one, two, five and ten years. These are not monitoring indicators.

The first four form one continuous paragraph on each page.

The question the Answer responds to is not written out, since it is implicit in the move from Complication to Answer. Missing information appears in grey and in braces, immediately after the statement that depends on it: information missing on .... Wherever the argument rests on something we do not know, the gap is marked rather than filled by inference.

The normative sources invoked are the Manual on the Organisation and Functioning of Health Posts and Health Centres of the National Health Service and the legal instruments it cites. References to the Manual give the section. The evidence for each action, and the diagnostic studies it depends on, are set out in the final section rather than on the action's page. The numbering of the domains reflects the order in which they were worked on, not an order of priority.

Domain 1 — Operating finance for first-level facilities

Presidential Decree 30/10 sets out the financing regime for municipal administrations and assigns them the provision of public goods and services in domains that include health, with budget priority given to the mandatory minimum expenditure of the health service; the Manual assigns the Ministry of Health methodological oversight of health planning and technical responsibility for service delivery (sections 3 and 4.3) Situation information missing on the full text of Presidential Decree 30/10 and Law 17/10, cited in the Manual and not read. The Manual itself records that this regime does not integrate the financial management of the municipality's health structures as a network, and that the result is a fragmentation of financial and management processes that does not favour efficient spending; it is not established whether any declared criterion governs the split between municipalities, nor what the difference is between the amount approved and the amount actually made available, so the Ministry answers for a performance whose means it does not control Complication information missing on the budget execution series by municipality and on what proportion of a facility's running cost flows through this route. None of the other actions reaches the facility if the money that keeps it running continues to be decided without a known rule, and the administrative reorganisation multiplies the units among which that money is divided Reason for sequencing. The investment action is to establish a formula for allocating primary health care operating finance based on population and need, executed by the municipal health authority and reported to the Ministry of Health Answer.

Lever 3 · Funding
The investment action

The investment action is to establish a formula for allocating primary health care operating finance based on population and need, executed by the municipal health authority and reported to the Ministry of Health.

Evidence and studies needed ↓
Possible actions orientations, not a recommendation · in order of ambition →
1
Transparency without changing the circuit

Publish the quarterly allocation and execution by municipality, leaving the decision where it is. Immediately feasible, almost no cost; does not correct the allocation but makes it visible.

2
Indicative formula

The Ministry of Health issues a formula based on population and need as guidance; the decision remains municipal. Requires no change in the law.

3
Earmarked allocation with a binding formula

A share of the municipal budget earmarked for primary health care, executed by the municipal health authority, accountable to the Ministry of Health. Requires a change in the law and agreement with Finance and with Territorial Administration.

Prior stepAll three require, as a prior step, calculating the running cost of a standard facility.
Actions to avoid
  • Recentralising the management of operating finance. It appears to correct the misalignment and lengthens the money's route.
  • Setting targets for primary health care spending before a budget classifier exists that identifies it. This produces an indicator that cannot be calculated.
  • Creating a dedicated vertical fund for primary health care. It solves the problem quickly and adds one more parallel circuit to those already in place.
Order of magnitude of cost

The action itself is analytical and normative: a unit cost study, the design of the formula and a publication mechanism. Low cost in absolute terms. What is at stake is not the cost of the action but the volume already circulating whose distribution is unknown. unit running cost of a first-level facility total operating finance for facilities, in aggregate and by municipality

Synergy with the other levers
Lever of the actionLevers activatedNo direct effect
Strategic
1Political commitment
2Governancemakes the division of competences verifiable
3FundingThis action
4Communities
Operational
5Models of care
6PHC workforcepredictability sustains rural incentives
7Physical infrastructure
8Medicineslast-mile delivery consumes operating resources
9Private sector
10Purchasing & paymentthe basis of any payment mechanism
11Digital technologies
12Quality
13Research
14Monitoringthe first comparable municipal series
What we expect to see a horizon, not monitoring indicators
1 year

running cost of a standard facility calculated; allocation and execution published by municipality.

2 years

formula in use, if only as guidance; the gap between approved and released amounts known.

5 years

dispersion between municipalities reduced; operating finance predictable at facility level.

10 years

needs-based allocation established as practice; the discussion moves to what is bought with the money.

Domain 2 — Empanelment of people to health teams

The Manual organises the territory into Health Zones of up to 7,500 inhabitants, assigned to Basic Health Teams that take health and social responsibility for that population, and provides that the empanelment of people to teams and facilities will be formalised progressively through the introduction of a Personal Card identifying users of the National Health Service (sections 4.3, 5.1 and 7.2) Situation information missing on whether the Personal Card was ever introduced in any province or municipality. The provision is written in the future tense and carries no deadline, no responsible body and no accompanying instrument; without formalised empanelment there is no list of people per team, coverage indicators including immunisation are calculated against denominators estimated from census projections that track neither people's changes of residence nor changes in administrative boundaries, and the system counts the contacts that took place without being able to name those who did not come Complication information missing on the census and projections in use, and on what the routine information system records at individual level. Naming those who did not come is the information missing from several problems currently treated separately, and the nominal register is the only element of population responsibility that survives a redrawing of administrative boundaries, because people stay the same when the polygons change Reason for sequencing. The investment action is to execute the empanelment already provided for in the Manual: a register of individuals and families linked to an identified Basic Health Team, with a person identifier, maintained by the team and reported in the routine information system Answer.

Lever 5 · Models of care
The investment action

The investment action is to execute the empanelment already provided for in the Manual: a register of individuals and families linked to an identified Basic Health Team, with a person identifier, maintained by the team and reported in the routine information system.

Evidence and studies needed ↓
Possible actions orientations, not a recommendation · in order of ambition →
1
Paper list at micro-area level

The community health agent keeps the list of families in the micro-area, which the Manual sets at an average of 150. Starts immediately with the arrangement already in place; ages quickly and cannot be aggregated.

2
Phased digital register by municipality

A nominal register linked to the routine information system, with a person identifier, rolled out in stages in selected municipalities.

3
National Personal Card linked to civil registration

This is the solution the Manual provides for. Greater reach and greater dependence on decisions outside the health sector.

Prior stepThe common prior step is deciding the person identifier. Choosing the medium before the identifier means doing the work twice.
Actions to avoid
  • Reopening the conceptual debate on whether there should be empanelment. It is settled in the norm; reopening it postpones execution.
  • Running a single complete enumeration. A register without continuous updating stops being useful within a few years.
  • Building a standalone register separate from the routine information system. It adds one more parallel system to those already running.
  • Treating population responsibility as discharged once the Health Zones are drawn. It is the most likely error, because it looks the same.
Order of magnitude of cost

Three items dominate: initial registration, technical support and continuous maintenance. Over ten years the third is the largest and is usually left out of the budget. The unit of effort follows from the norm of 150 families per micro-area. cost of registration per family number of community health agents actually deployed cost of the person identifier and of the digital support

Synergy with the other levers
Lever of the actionLevers activatedNo direct effect
Strategic
1Political commitment
2Governance
3Funding
4Communitiesthe agent answers for a list rather than an area
Operational
5Models of careThis action
6PHC workforcedefines each team's workload
7Physical infrastructure
8Medicines
9Private sector
10Purchasing & paymenta defined population makes capitation possible
11Digital technologiesgives digitisation a purpose
12Quality
13Research
14Monitoringsupplies the denominator for coverage
What we expect to see a horizon, not monitoring indicators
1 year

identifier defined; registration instrument approved; first municipalities with nominal lists.

2 years

share of the population with an identified team known and rising.

5 years

coverage calculated against the empanelled population; possible to name who is missing and go and find them.

10 years

longitudinal follow-up of the same person across contacts and across levels of care.

Domain 3 — Clinical regulation between levels of care

The Manual establishes referral and counter-referral pathways between Health Posts, Health Centres and Municipal Hospitals, and assigns the Ministry of Health the normative function of developing, for each type of service and service package, protocols and, where appropriate, practical guidance, to be made available through the public health and facility management programmes (sections 4.6, 7.2 and 7.6) Situation information missing on which protocols have actually been produced since then, by whom, with what status and whether they are in force. The provision is written in the future tense and carries no deadline, no named responsible body, no revision cycle and no route to publication; the development of the referral system is delegated to a service at facility level, and the Manual goes no further than stating that it should take account of what is referred, how it is referred and the conditions for referral, without establishing any criterion; it is therefore not possible to say whether a referral was appropriate, nor does the receiving facility have any basis for deciding what to send back and to whom Complication information missing on whether a standard counter-referral instrument exists and whether completing it is required. The referral criterion is what makes the Common Service Package, the staffing norms and the essential medicines list by level applicable, and it is the only benchmark against which the first level's capacity to resolve cases can be assessed Reason for sequencing. The investment action is to execute that normative function: national referral criteria by clinical condition and by level of care, with institutional authorship, normative status, revision cycle and route to publication defined, and with a mandatory counter-referral instrument Answer.

Lever 12 · Quality
The investment action

The investment action is to execute that normative function: national referral criteria by clinical condition and by level of care, with institutional authorship, normative status, revision cycle and route to publication defined, and with a mandatory counter-referral instrument.

Evidence and studies needed ↓
Possible actions orientations, not a recommendation · in order of ambition →
1
Adopt and adapt for a small set of conditions

Start from the algorithms already in use in the vertical programmes and from international references, for the highest-volume conditions. Fast and cheap; partial coverage.

2
Produce criteria for the whole Common Service Package

Complete coverage; requires clinical capacity and time, and produces a document that ages if no one maintains it.

3
Create a permanent clinical regulation function

A body with a mandate to produce, publish and revise the criteria on a defined cycle. The only one of the three that solves the problem of currency.

Actions to avoid
  • Setting referral completion targets before the criterion exists. Without a criterion, the rate rises if fewer or easier cases are referred.
  • Producing a single long manual. What gets used during a consultation is what is short and to hand.
  • Leaving each vertical programme with its own algorithm and no articulation between them. It multiplies instructions that meet in the same patient.
  • Making counter-referral compulsory without ensuring someone reads it and acts on it. It produces paper, not continuity.
Order of magnitude of cost

Cost dominated by specialist technical time rather than procurement: clinical panels, drafting, validation, publication and periodic revision. Probably the cheapest of the four actions and the one that consumes the scarcest resource, namely professionals with clinical competence and protected time for this work. number of conditions to cover whether national capacity exists to convene clinical panels cost of getting the criterion to the point of consultation

Synergy with the other levers
Lever of the actionLevers activatedNo direct effect
Strategic
1Political commitment
2Governance
3Funding
4Communities
Operational
5Models of caremakes the Common Service Package applicable
6PHC workforcethe basis of professional profiles and training
7Physical infrastructure
8Medicinesdefines what must exist at each level
9Private sector
10Purchasing & payment
11Digital technologiesthe content the digital tool carries
12QualityThis action
13Research
14Monitoringmakes capacity to resolve measurable
What we expect to see a horizon, not monitoring indicators
1 year

criteria for a small set of conditions published, with authorship, status and revision date.

2 years

counter-referral instrument in use; appropriate and inappropriate referrals distinguishable in a sample.

5 years

first-level capacity to resolve cases measured against the criterion, by municipality.

10 years

clinical regulation function operating permanently, with criteria that track changes in the Package.

Domain 4 — Continuing education tied to clinical practice

A national training programme for 38,000 health professionals is under way, and the Manual states that the system should move towards integrated and formative supervision, tasking the Provincial and Municipal Health Directorates with establishing their supervision protocols and an annual plan (section 8.4) Situation information missing on the scope, content, duration and state of execution of the training programme. Formative supervision is stated as a direction of travel rather than as a norm: the Manual defines no content, no frequency, no instrument and no account of who observes what, and it contains no reference at all to continuing education; training is therefore counted by professionals trained rather than by practice changed, with no knowledge of whether those trained work where the competence is needed, whether they apply it, or whether the facility's performance changed afterwards Complication information missing on whether a training record exists per professional, on the frequency of supervision and on who carries it out. A revised service package, a referral criterion or a task-sharing arrangement only reach the patient if someone changes what happens in the consultation, which makes this the action that carries the content of all the others to the point of delivery Reason for sequencing. The investment action is to establish a continuing education cycle anchored in the facility's own practice — formative supervision with observation of work, review of cases and of the team's own data, and a training record per professional — in place of one-off training events Answer.

Lever 6 · PHC workforce
The investment action

The investment action is to establish a continuing education cycle anchored in the facility's own practice — formative supervision with observation of work, review of cases and of the team's own data, and a training record per professional — in place of one-off training events.

Evidence and studies needed ↓
Possible actions orientations, not a recommendation · in order of ambition →
1
Turn the supervision visit into a teaching act

A single instrument, observation of work and feedback to the professional. Uses the structure the Manual already provides for; depends on the supervisor having the clinical competence to teach.

2
Regular clinical session in the facility

The team reviews its own cases and its own data at a fixed frequency. Almost the only cost is protected time; requires someone able to lead it.

3
Accredited programme with a record per professional linked to career progression

Greater reach and permanence; requires a regulatory decision on careers and a body to accredit.

Prior stepThe common prior step is having a criterion of correct practice to train against. Without one, supervision observes without knowing what it is looking for.
Actions to avoid
  • Measuring continuing education by the number of professionals who attended training. It rises with the supply of events, not with changed practice.
  • Cascade training without verifying what reaches the end of the chain.
  • Taking the professional out of the facility for training without arranging cover. It withdraws service from the population one is trying to serve.
  • Training competences the facility has no means to exercise, for want of medicines, equipment or time.
  • Using the supervision visit as an inspection. If it produces a report and gives nothing back to the professional, it does not teach.
Order of magnitude of cost

A recurrent rather than capital cost, dominated by professionals' time: the supervisor's, the team's protected time, and cover for those withdrawn from service. The cheapest option is the in-facility session, which has almost no monetary cost and a high organisational one. The most expensive element is not the training but the travel. cost per professional trained in the national programme number of supervisors and current frequency of visits cost of cover during training

Synergy with the other levers
Lever of the actionLevers activatedNo direct effect
Strategic
1Political commitment
2Governance
3Funding
4Communitiescommunity health agents join the same cycle
Operational
5Models of carecompetences determine what the facility can deliver
6PHC workforceThis action
7Physical infrastructure
8Medicinestraining prescribers changes prescribing patterns
9Private sector
10Purchasing & payment
11Digital technologies
12Qualitythe vehicle for clinical audit
13Research
14Monitoringthe record per professional creates the baseline
What we expect to see a horizon, not monitoring indicators
1 year

single formative supervision instrument in use; training record per professional started.

2 years

frequency of supervision known and rising; regular clinical sessions in selected municipalities.

5 years

measurable change of practice in a chosen domain, for example prescribing, rather than numbers trained.

10 years

continuing education tied to career and accreditation, with its own revision cycle.

Domain 6 — National digital health governance framework

The Manual identifies the clinical record and clinical documentation as the main information system of the Health Post, and assigns to the Admissions and Clinical-Statistical Records Service the management of information systems, data collection, statistics and databases (section 7.2); it contains no reference to digital systems of any kind Situation information missing on whether any national eHealth or digital health strategy has been adopted, and its date. In practice records are kept on paper, programme by programme. At the same time three digital systems are being scaled up in parallel — DHIS2 as the national platform, REDIVE and IOTA at community level — with no instrument stating which is the system of record, what each is for, how they exchange data, or who governs them; the draft strategy itself records the absence of a national digital health governance framework Complication information missing on who commissions and funds REDIVE and IOTA, what each records, and at what level. Every other digital action in the set — the person identifier, the clinical record, the digital regulation of referral, teleconsultation — builds on a prior decision about architecture and data standards; taken without it, each will produce its own system and the parallel systems already visible will multiply Reason for sequencing. The investment action is to establish a national digital health governance framework that designates the system of record for each type of data, sets data standards and exchange rules, assigns institutional responsibility, and carries the telehealth and data protection provisions the other actions depend on Answer.

Lever 11 · Digital technologies
The investment action

The investment action is to establish a national digital health governance framework that designates the system of record for each type of data, sets data standards and exchange rules, assigns institutional responsibility, and carries the telehealth and data protection provisions the other actions depend on.

Evidence and studies needed ↓
Possible actions orientations, not a recommendation · in order of ambition →
1
Designation decision

A ministerial decision naming the system of record for routine aggregate data, for community data and for individual data, without a full strategy. Settles the immediate conflict quickly; leaves standards open.

2
National digital health strategy with architecture

Architecture, data standards, interoperability, telehealth policy, data security and system specifications — the content the PHCMFI indicator 42 requires. Comprehensive; slower.

3
Governance body with a standing mandate

A unit with authority to approve any new digital system before deployment, maintain the standards and revise the architecture. The only one of the three that stops the problem from recurring.

Prior stepThe common prior step is an inventory of the systems in operation: what exists, who funds it, what it records, and who uses it.
Actions to avoid
  • Commissioning a new platform to replace the existing ones before deciding what each is for. It adds a fourth system to the three already running.
  • Letting each donor-funded programme choose its own tool. This is how the current parallel systems arose.
  • Writing the framework as a technology document. The decisions it must settle are institutional: who owns the data, who approves a system, who pays for its maintenance.
  • Digitising before there is an individual record worth digitising. A digital system reproduces the process it replaces; where there is no individual record on paper, there is nothing to carry over.
Order of magnitude of cost

Low as a normative product: drafting, consultation and approval. The cost it governs is much larger — every digital system approved or rejected under it. The main recurrent item is the staff time of whoever holds the governance function. number and cost of digital systems currently funded in the health sector, by funder existence of an institutional home with capacity to host the governance function

Synergy with the other levers
Lever of the actionLevers activatedNo direct effect
Strategic
1Political commitment
2Governanceassigns institutional responsibility for data
3Funding
4Communitiesbrings community reporting into the same architecture
Operational
5Models of caredefines what the clinical record carries
6PHC workforce
7Physical infrastructure
8Medicinesconnects supply chain systems to facility data
9Private sector
10Purchasing & payment
11Digital technologiesThis action
12Quality
13Research
14Monitoringone system of record ends parallel reporting
What we expect to see a horizon, not monitoring indicators
1 year

inventory of systems in operation completed; system of record designated for each type of data.

2 years

framework approved with data standards and exchange rules; every new system subject to prior approval.

5 years

parallel systems reduced; community, supply chain and facility data exchanged through defined interfaces.

10 years

an architecture that new tools join rather than replace.

Evidence

The basis on which each action rests, with the status of each element marked in braces.

Domain 1 — Operating finance for first-level facilities

Basis
  • Manual, sections 3 and 4.3read
  • Presidential Decree 30/10 and Law 17/10cited in the Manual; full text not read
  • Diagnosis of financial fragmentationstated by the Manual itself, section 3
  • Budget execution disaggregated by municipalitynot available in the material at hand
  • Unit running cost of a first-level facilitydoes not exist
Diagnostic studies needed
Financing circuit and public financial management review

how money reaches a facility, and the difference between approved, authorised and executed amounts by municipality

Health accounts (SHA 2011)

current and PHC-specific expenditure, which cannot be derived from the 2008 accounts

Costing study

the unit running cost of a standard facility, without which the formula cannot be calibrated

↑ Back to Domain 1

Domain 2 — Empanelment of people to health teams

Basis
  • Manual, sections 4.3, 5.1 and 7.2read
  • Introduction of the Personal Cardnot established
  • Annex 3 of the Manual, Practical Guide to the Admissions and Clinical-Statistical Records Servicereferred to in the Manual; not available
  • Number of community health agents deployeddoes not exist
  • Unique person identifierexistence not established
  • Capacity of the routine information system to record at individual levelunknown
Diagnostic studies needed
Population survey module

the proportion of people empanelled to an identified team, which can only be measured in households, not in facilities

HHFA

whether facilities keep any list of users or family records

Digital landscape assessment

what the routine information system records at individual level, and whether a person identifier exists

↑ Back to Domain 2

Domain 3 — Clinical regulation between levels of care

Basis
  • Manual, sections 4.6, 7.2 and 7.6read
  • Protocols and guidance produced under section 7.6existence not established
  • Clinical algorithms of the vertical programmeslikely to exist, content unknown
  • Current volume and pattern of referralsdoes not exist
  • Share of patients accessing hospital level directlydoes not exist
Diagnostic studies needed
Legal and normative review

which protocols, if any, have been produced since 2017, by whom and with what status

Routine data review

whether referrals are recorded anywhere and what can be learned from existing registers

HHFA

whether any referral or counter-referral instrument exists in facilities

↑ Back to Domain 3

Domain 4 — Continuing education tied to clinical practice

Basis
  • Manual, section 8.4read; states a direction of travel, with no operational content
  • National training programme for 38,000 professionalsscope, content and execution unknown
  • Supervision protocols of the Provincial and Municipal Health Directoratesexistence not established
  • Training record per professionalexistence not established
  • Effect of training on clinical practice in Angoladoes not exist
Diagnostic studies needed
Health workforce accounts

staff numbers by category and facility, and whether a training record per professional exists

HHFA

whether supervision takes place, how often, and what is observed

↑ Back to Domain 4

Domain 6 — National digital health governance framework

Basis
  • Manual, section 7.2read; describes a paper-based information system with no digital provision
  • Absence of a national digital health governance frameworkrecorded in the consultants' draft, p. 38
  • REDIVE, IOTA and DHIS2 scaling in parallelrecorded in the draft, p. 23; ownership, scope and funding unknown
  • Records kept on paper in facilitiesfield observation, not documented
  • National eHealth strategyexistence not established
  • Required content of a digital health strategyPHCMFI indicator 42
Diagnostic studies needed
Digital landscape assessment

the inventory of systems in operation, who funds them, what they record and how they exchange data

Legal and normative review

whether a national eHealth strategy or any data protection provision exists

↑ Back to Domain 6

The diagnostic studies

The studies named above are drawn from a set of twelve identified across the full list of investment actions. They are what would have to be known before these actions can be dimensioned. Two of them serve most of the set: the Harmonized Health Facility Assessment, which shows what exists in facilities, and the legal and normative review, which shows what exists in the instruments. They are complementary rather than overlapping, and neither can answer the other’s questions — facility instruments give no coverage of governance and financing, which can only be reached through document review and interviews.

The full mapping of studies to actions, with the remaining studies not cited here, is kept in the working spreadsheet of aggregated actions.

Note on the nature of the missing information

The braces marked across the five domains are not all of the same kind.

1

Some can be resolved by reading documents that exist and that we can obtain, notably the legal instruments cited in the Manual and its annexes.

2

Others can be resolved by requesting data from whoever holds it, in particular budget execution series by municipality and referral records.

3

A third group cannot be resolved by reading or by data, because it concerns normative products the Manual requires to be produced and which may never have been produced; in those cases the answer can only be obtained from those with the authority to produce them.