Designing anticholinergic burden research in schizophrenia

Abstract

Anticholinergic-burden research in schizophrenia requires a clear distinction between medication classification, actual exposure, cognitive performance, and treatment benefit. This article proposes a study-design framework grounded in the cited schizophrenia literature and UnifiedACB’s source-specific evidence model. It is not a registered study or a report of newly collected participant data. The proposed methods define a target population and care phase, freeze the source version and identity rules, establish an exposure window, and select a primary outcome before analysis. They also require an account of treatment indication, co-medication, incomplete histories, and repeated testing. No sample size, effect estimate, or clinical validation result is asserted. The expected deliverable is a reproducible protocol whose findings could be interpreted within their actual population and measurement limits. The framework does not endorse a universal score threshold or use a synthetic average of ordinal scales.

Introduction

Schizophrenia studies can differ substantially in illness duration, stability, treatment history, care setting, and cognitive assessment. These differences can change both who receives classified medicines and how outcomes are measured. A study cannot be made reproducible merely by naming an anticholinergic scale.

The objective is to outline decisions that should be resolved before collecting or analyzing data. The framework can inform a future observational study or, with appropriate additional specifications, a study of a medication-review strategy. Those designs should remain distinct because their causal interpretations are different.

Methods

The proposed protocol would first define the population and recruitment setting. It should specify diagnostic eligibility, age range, care phase, and any stability requirement, with a rationale tied to the research question. First-episode care, long-term outpatient treatment, and acute inpatient care should not be pooled without considering whether the exposure and outcome windows mean the same thing.

The exposure specification would identify the original source and version, the medication-history source, and the period represented. It should distinguish active orders, dispensing, reported use, and recorded administration. Scheduled and occasional medicines need explicit handling. Exact ingredient, formulation, route, and combination-product rules should be documented before matching records.

The source observations would remain immutable. A published zero would be retained as zero; an absent or unresolved observation would remain incomplete. The protocol should report coverage for the full regimen and specify how incomplete records enter each analysis. A second source may be used for a planned sensitivity analysis, but its values should not be averaged into the first source’s categories.

The outcome plan would identify one primary endpoint and a clinically justified assessment interval. If cognition is primary, the instrument, domain or composite, scoring direction, and conditions of repeat testing should be specified. Other outcomes, such as symptom control, treatment tolerability, or functional participation, should be labeled secondary and measured on their own terms.

The analysis plan would state whether the goal is descriptive association or evaluation of an intervention. It should identify plausible baseline confounders from the clinical question, distinguish them from post-treatment variables, and explain model choices. Sample size would need a separate calculation using the chosen outcome and realistic data assumptions; no number is supplied here without that information.

Finally, the protocol would define data-quality checks, missing-data reporting, participant flow, and deviations from the planned analysis. If an intervention is studied, the treatment strategy, permissible individualization, clinical oversight, and monitoring of the treated illness must be specified. The evidence directory cannot supply those operational or ethical requirements by itself.

Results

No participants have been recruited and no new outcome analysis is reported in this article. The result is a proposed specification for future work, not evidence that a burden threshold or medication-review strategy has been validated.

Existing schizophrenia research provides a rationale for asking the question. Georgiou and colleagues’ review emphasizes heterogeneity in studies relating exposure to cognition. Georgiou et al., 2021 That observation supports prespecifying population and measurement choices. It does not supply the missing protocol decisions, a guaranteed effect size, or a justification for treating all study settings as equivalent.

Discussion

A transparent protocol should make it possible to reconstruct not only the final source total but also the medicines and decisions behind it. If two participants have the same total with different contributors and coverage, that distinction may matter to interpretation. Contributor-level reporting can reveal whether a result is driven by a particular treatment pattern or by how missing observations were handled.

The care phase is equally important. An exposure recorded during acute treatment may not represent the interval over which a later cognitive outcome developed. A stable outpatient regimen offers a different design opportunity but may introduce selection related to who remains in follow-up. Neither setting is automatically superior; it must fit the stated question.

Outcome interpretation should remain clinically bounded. A statistically detectable difference in a test score is not necessarily a meaningful change in daily life. Conversely, a study may be unable to detect a clinically relevant effect if measurement is imprecise or follow-up incomplete. The report should describe uncertainty rather than equate a nonsignificant result with proof of no effect.

The protocol should also prevent selective source choice. If investigators compare several scales, report the prespecified role of each and the corresponding coverage. Different findings can be informative about the methods rather than an inconvenience to be hidden. Publication should retain the original source values and clearly separate any analytic transformations.

Record clinical changes that can alter interpretation

A prospective study should document relevant treatment and clinical changes during follow-up, with dates sufficient to understand their relationship to cognitive assessment. A new adjunct, a change in actual as-needed use, or a change in illness state may make the baseline regimen an incomplete description of later exposure. The protocol would need to decide which changes belong in the primary analysis and which warrant sensitivity analyses, before their observed association determines that choice.

If the study examines a supervised medication change, the reason for that change should be retained alongside the outcome. A change initiated because of a cognitive complaint differs from one prompted by movement symptoms, treatment failure, or another clinical need. Combining these pathways without describing them can obscure what a subsequent improvement or deterioration means. No such intervention or follow-up has been performed for this proposed protocol.

Limitations

This framework is not an operational protocol ready for recruitment. It does not specify a dataset, institution, sample size, consent process, randomization procedure, or statistical model fitted to data. Those elements require a concrete study question and appropriate review. No ethics approval or independent validation is implied.

Medication histories may be incomplete, treatment selection may be confounded, and cognitive testing may be affected by changing clinical conditions. Even a well-executed study in schizophrenia would not automatically validate a threshold in mood disorders, children, or another health system. The source library provides provenance, not universal clinical transportability.

Conclusion

A credible schizophrenia study begins with a defined population, exposure window, outcome, and analysis purpose. Freezing the source version and preserving missingness makes the evidence reproducible; a well-specified clinical design makes the resulting conclusion interpretable. Both are needed before claiming validation or recommending a burden-based treatment target.

References

1. Mancini V, et al. Anticholinergic Burden and Cognitive Function in Psychosis: A Systematic Review and Meta-Analysis. Am J Psychiatry. 2025;182:349-359. Source
2. Vennard O, Stewart C, Tolia M, Soiza RL, Myint PK. Anticholinergic Medication Burden Scales: A Systematic Review. J Am Geriatr Soc. 2026;74(6):1771-1784. Source
3. Georgiou R, et al. Anticholinergic Burden and Cognitive Performance in Patients With Schizophrenia: A Systematic Literature Review. Front Psychiatry. 2021;12:779607. Source

Disclosure

Funding: None. Conflict of interest: None declared for the UnifiedACB educational project.

Educational evidence summary. This content does not provide patient-specific medical advice and does not replace clinical judgment, medication reconciliation, or prescribing guidance.