HEOR Systematic Literature Review Protocol Template

A decision-ready HEOR or HTA review needs more than a conventional systematic-review protocol. It must preserve the connection between the institutional decision, the relevant comparator set, the populations that matter to that decision, the evidence streams needed to support it, and the timing at which the evidence package will be judged. NICE explicitly organizes evidence submissions around a decision problem and relevant comparators; PBAC begins clinical evaluation from the proposed indication and nominated comparator; current EU joint clinical assessment uses a PICO-defined assessment scope that may contain several decision questions.^1,2,7,13

This template is designed for that planning layer. It does not reproduce PRISMA-P fields, Boolean search syntax, screening mechanics, risk-of-bias tutorials or meta-analysis instructions. Those remain important, but they are generic review infrastructure rather than the distinctive problem this page solves. Here, the task is to configure an evidence program that remains usable for a payer, HTA body, comparative-effectiveness team or economic model when the review moves beyond publication and into a decision process.^15,17,18

What this protocol adds beyond a generic systematic-review protocol

A generic protocol can be methodologically strong and still produce evidence that is poorly suited to HTA. The failure usually occurs before screening. The review may include the population used in pivotal trials while the reimbursement decision concerns a narrower treatment line. It may compare the technology only with placebo even though the relevant payer comparator is active treatment. It may collect efficacy endpoints but omit treatment switching, subsequent therapy, candidate effect modifiers or data-cut maturity that later become necessary for comparative-effectiveness work. It may finish six months before submission without any rule for handling a new pivotal readout.

Before you fill the template: define the evidence program, not just the review

Start by asking which evidence functions the decision actually needs. A clinical-effectiveness stream is common in reimbursement work, but epidemiology, natural history, treatment patterns, RWE, health-state utilities, resource use or published economic evaluations are conditional. NICE separates several of these evidence classes in company submissions, and Cochrane’s economics guidance similarly distinguishes full economic evidence review from lighter or absent economic components depending on the decision and available resources.^1,16

The protocol should therefore assign each stream a purpose. “Utilities” is not a purpose by itself. The team should know whether utility evidence is needed to populate a particular health state, test the applicability of trial-derived EQ-5D values, or identify a mapping source. “RWE” is equally incomplete unless the intended use is stated: external control, natural history, treatment patterns, safety, resource use or contextualization. NICE’s RWE framework is explicitly use-case oriented, and that logic is more useful than treating observational evidence as one broad category.⁵

Build the HEOR/HTA protocol blueprint

The builder below is intentionally modular. It is not a scoring tool. A blank field is not a methodological failure; it is a prompt to decide whether the information is known, genuinely not applicable, or unresolved and therefore governed by an amendment or feasibility rule. Entries are stored only in the current browser through local storage.

Complete the relevant modules, then select “Generate protocol blueprint”.

A master evidence universe can serve several PICOs—if the mapping stays explicit

Multi-PICO work creates a common design error: teams either duplicate the entire search for every small variation or, at the opposite extreme, treat several decision questions as though they were one. The safer architecture sits between those extremes. When the disease, intervention and broad evidence universe substantially overlap, one coordinated identification strategy may serve several PICOs. Eligibility mapping, comparator relevance, subgroup definitions and analysis still need to remain PICO-specific. PBAC similarly distinguishes searches needed to identify direct evidence and comparator trials for indirect comparison, while EU JCA operational experience shows that missing evidence for some PICO questions can leave large parts of an assessment scope unanswered.^7,8,13,14

Do not interpret this as a rule that all PICOs should share one search. Separate strategies become more defensible when the evidence streams use different study designs and terminology, when the intervention or disease area changes materially, or when an economic/utility question has a different source architecture from the clinical-effectiveness question. The protocol should justify the architecture rather than use a fixed number of searches as a marker of rigor.

Treat a trial as an evolving evidence object

Study lineage is one of the most useful additions a decision-oriented protocol can make. Cochrane requires review authors to link multiple reports of the same study and to plan how discrepancies across sources will be handled. PBAC goes further operationally by asking for a master list of trials and associated reports. In HTA work, the same study may appear as a registry record, conference abstract, primary paper, CSR, regulatory assessment and later survival update. Counting those records as separate evidence is an obvious error, but silently replacing one with another is also risky.^14,15

The protocol should not use a crude hierarchy such as “peer-reviewed article always wins” or “latest source always wins.” Different sources can be strongest for different variables. A later data cut can provide more mature survival while the original CSR may contain fuller safety or protocol information. The safest general rule is variable-specific reconciliation: retain a master study ID, record each source and data-cut date, mark supersession status, document the preferred source for each variable, and preserve unresolved conflicts. This is an Academy synthesis of well-established study-linkage principles rather than a claim that every HTA body mandates an identical hierarchy.

Evidence currency is a governance problem, not a calendar slogan

A review can be complete and still be stale for its decision. The protocol should distinguish at least four dates: search execution, evidence cutoff, analysis/data freeze, and target submission or assessment milestone. They answer different questions. Search execution documents when a source was queried. Evidence cutoff defines which new evidence can enter the current analytical dataset. Data freeze identifies the trial-data maturity used for outcomes. Submission timing determines whether surveillance or a rerun may be necessary before the evidence package is locked.

Do not invent a universal “update within three months” or “six months” rule. HTA agencies use different procedural cutoffs, and a high-velocity oncology evidence base has a different risk profile from a stable utility search. The protocol should instead define material triggers: a new pivotal trial, a new relevant comparator, a revised scope, an abstract-to-full-publication transition, a new data cut, a regulatory change, a safety signal, a delayed submission or a resubmission. Cochrane’s review-update standards demonstrate that currency matters, but their own timing requirements should not be presented as universal HTA rules. If a project is deliberately maintained as a living systematic review, PRISMA-LSR can guide transparent reporting of that living process; it does not create a universal HTA update cadence.^19,22

Use proportional search depth for economic inputs

An HTA evidence program can waste substantial effort by treating every model parameter as if it requires a full systematic review. NICE DSU TSD 13 addresses the identification and selection of evidence used to inform model structure and parameters, while Cochrane’s economics guidance explicitly recognizes different levels of economic-evidence incorporation. The practical implication is not a numeric scoring algorithm. It is a transparent decision about search depth based on the decision use of the parameter.^3,16

Worked planning example: an oncology reimbursement evidence program

Suppose a new oncology medicine has a pivotal single active-control trial, but the target payer considers two additional active therapies relevant in later-line practice. The economic model also needs health-state utilities and local resource use. A generic protocol might define the pivotal-trial PICO, search MEDLINE and Embase for the new medicine, and plan meta-analysis “if possible.” That protocol could be publishable and still fail operationally.

A decision-linked protocol would begin by recording the reimbursement population, treatment line and all decision comparators. It would then create a clinical master evidence universe broad enough to identify trials of the new medicine and relevant comparator evidence, while assigning each comparison a stable PICO mapping. The clinical stream would prospectively collect trial arms, comparator identity, baseline characteristics, outcomes, timepoints, switching/subsequent therapy and data-cut/source information. It would mark IPD and aggregate-data availability without assuming that an ITC or NMA will later be valid. Pair 31 and Pair 32 methodology would be applied only after the evidence structure is known.

The utility question would be a separate stream because its eligible study designs, sources and extraction fields differ. The resource-use question might rely partly on targeted literature and partly on authoritative local costing sources. Every record from the pivotal trial would be linked to one master study ID so that a conference readout, journal article and later regulatory data cut do not become three “studies.” The update plan would specify what happens if a new comparator trial or mature survival cut appears before submission. If the payer changes the required comparator, the protocol amendment would identify which searches, PICO mappings, extraction fields and downstream feasibility checks need to be reopened.

What this planning resource does not decide

This resource does not replace a complete systematic-review protocol, a search-strategy protocol, a jurisdictional dossier template, a risk-of-bias manual, or an economic model specification. PRISMA-P remains reporting guidance for systematic-review protocols; PRISMA-S covers search reporting; PROSPERO and OSF provide registration infrastructure with different scopes. The builder should sit alongside those requirements rather than pretending to supersede them.^17,18,20,21

It also deliberately stops before specialist quantitative judgments. Capturing candidate effect modifiers is different from deciding whether transitivity holds. Recording a common comparator is different from deciding that an indirect comparison is credible. Capturing economic-input evidence is different from choosing a model structure. Those boundaries protect the Academy’s separate methodology pages from becoming keyword-swapped versions of one another and keep this resource focused on a single task: making the SLR protocol usable for a real HTA/HEOR evidence program.

Related methodology

A protocol is decision-ready when it can explain what happens next

The practical test is not whether the protocol has more headings. It is whether the evidence team can trace each major decision requirement to an evidence stream, each PICO to its comparator logic, each study to its underlying record family and data cut, each planned downstream analysis to the variables it will require, and each update or amendment to a defined governance rule. That is the difference between adding “HEOR” to a conventional protocol and designing an evidence program that can survive an HTA decision.