Diary-Based Contact Surveys Get a Standardized Core After Pandemic-Driven Fragmentation
Diary-based social contact surveys sit at the base of many transmission models for close-contact infections. When the contact diary is inconsistent, the model inherits the inconsistency. During COVID-19, that weakness became impossible to ignore: survey teams were collecting rapid, repeated data under pressure, but they were doing so with different contact definitions, different recall periods, different sampling strategies, and different questionnaire structures. The result was a crowded literature that was rich in volume and uneven in comparability.
This systematic review assembled that literature and turned it into something more usable. Across 129 studies from 62 countries, the investigators mapped study design, survey methodology, contact definitions, and questionnaire content from diary-based social contact surveys published between 31 January 2018 and 25 September 2025. The pattern was stark. Most studies clustered in 2021 and 2022, a period when public health teams needed behavioral data fast and often had to build instruments on the fly.
One finding captures the methodological drift of the pandemic era: retrospective reporting dominated, and non-probability sampling became routine.
A Pandemic Literature Heavy on Speed, Light on Harmonization
Retrospective reporting appeared in 86.0% of studies, while non-probability sampling was used in 77.5%. Both approaches became increasingly common during the pandemic, and non-paper-based data collection rose sharply from 2021 onward. Those choices are understandable in an emergency. Paper diaries, in-person recruitment, and probability-based fieldwork are slower, costlier, and often harder to sustain during mobility restrictions or repeated survey waves. Yet each shortcut also changes what the data can and cannot claim about population mixing.
For transmission modeling, those design decisions matter because close-contact infection risk is not shaped by contact counts alone. It is shaped by who meets whom, where those encounters occur, how long they last, and whether they involve physical proximity or shared indoor air. A survey that captures contacts poorly can distort age-specific mixing matrices, household transmission estimates, workplace exposure patterns, and any model built downstream. The review therefore treated questionnaire design not as administrative detail, but as a determinant of model validity.
The review also assessed methodological quality and reporting with an adapted AXIS tool, summarizing results by criterion. Even without a single pooled effect estimate, the message is clear: the field has generated a large evidence base, but the reporting conventions remain too heterogeneous for straightforward comparison across settings and time periods.
Fifty-Four Survey Items Narrowed to Twenty-Two Core Elements
The questionnaire work gives the review its most practical output. Among 36 retrievable questionnaires from general-population surveys, the investigators identified 54 distinct survey items. An online assessment with 26 experts then classified those items by perceived importance, producing an expert-informed questionnaire with 22 core items.
This narrowing process is more than an editing exercise. Contact surveys can become bloated with redundant questions that burden participants, reduce completion quality, and make repeated deployment harder during outbreaks. At the same time, stripping a survey too far can erase the details needed to understand transmission. The value of the 22-item core lies in balancing those two pressures: enough information to support model fitting and intervention analysis, but not so much complexity that the instrument collapses under emergency use.
The resulting core elements are intended to improve harmonization across studies rather than impose a one-size-fits-all survey for every pathogen or setting. Close-contact infections vary in their mode of spread, generation time, and susceptibility to intervention, and a useful questionnaire has to preserve enough flexibility to reflect that biology. Even so, a shared backbone allows studies to be compared across countries, pandemic phases, and survey platforms with far less ambiguity than the current literature permits.
Updated Guidance Targets Sampling, Definitions, and Data Governance
Based on the review findings, the investigators updated previous recommendations on sampling, survey methods, contact definitions, behavioural change, essential contact information, and data governance. Each of those domains has a direct effect on inference. Sampling strategy shapes representativeness. Contact definitions determine whether an encounter is counted at all. Survey mode influences recall and response burden. Behavioral-change items matter because contact patterns can shift abruptly during outbreaks in response to risk perception, policy, school closures, workplace disruption, or isolation guidance.
Data governance may sound administrative, but in outbreak research it is part of the scientific infrastructure. Contact surveys often contain sensitive details about daily routines, household composition, workplace attendance, and social behavior. Clear governance standards help preserve participant trust while also supporting timely data sharing across research teams and public health agencies.
The recommendation set reflects a field that has moved from exploratory measurement toward operational epidemiology. The pandemic showed how urgently models need timely contact data, but it also exposed how fragile those data become when every group designs its own instrument from scratch. A harmonized survey framework does not eliminate local adaptation; it makes adaptation legible.
Why Contact Diaries Will Matter in the Next Outbreak Window
Close-contact transmission models are only as credible as the contact data feeding them. When a respiratory pathogen emerges, or when a known infection starts to spread in a new way, public health teams need to know how exposure is distributed across age groups, settings, and daily routines. Diary-based surveys remain one of the few tools that can capture that information at scale in near real time. Their utility, however, depends on whether the underlying questions are stable enough to compare across waves and flexible enough to deploy under pressure.
This review does not solve every measurement problem, and it does not convert every historical survey into a directly comparable dataset. It does something more concrete: it sets a sharper reference point for future contact studies and offers an expert-informed core questionnaire that researchers can build on rather than reinventing the wheel for each outbreak. The article also makes plain that the largest surge in survey production came during COVID-19, when urgency outpaced standardization and retrospective, non-probability designs became the norm.
For future preparedness, that imbalance is the central lesson. The next emergency will again demand fast behavioral data, but the value of that data will rise or fall on whether investigators use a common language for defining contacts, recording behavior change, and documenting the circumstances under which exposure is measured.