JMIR Publications

ABSTRACT

Background:

The COVID-19 pandemic has triggered major global challenges, highlighting the need for timely surveillance and accurate forecasting to curb disease transmission and reduce the health burden. Traditional modelling approaches usually rely on aggregated data on confirmed cases. However, testing data can be unreliable due to low availability and uptake as well as registration delays. In Sweden, the 1177 healthcare helpline provides valuable syndromic surveillance data that can capture early signals of infection prior to confirmation through testing.

Objective:

This study aims to assess the effect of integrating syndromic surveillance data on short-term forecasting accuracy by evaluating its impact on the predictive performance of epidemiological models for COVID-19.

Methods:

We developed a novel epidemiological model named SI_s I_c HRD, incorporating data from the 1177 healthcare helpline cross-linked with national registers on COVID-19 testing, hospitalisations, and deaths. In the novel model the infected population is divided into suspected infections (I_s; those reporting symptoms to 1177) and confirmed infections (I_c; those who tested positive) individuals. Accuracy metrics of this model are compared to a model without data from the 1177 healthcare helpline. To evaluate model performance across different pandemic phases between 2 March and 31 December 2020, we used a daily rolling-window approach with a 60-day fitting period to predict confirmed infections, hospitalisations, and deaths over the subsequent 30 days. Prediction performance was evaluated using Mean Absolute Error (MAE) and Mean Absolute Percentage Error (MAPE).

Results:

The proposed SI_s I_c HRD model successfully captured the temporal dynamics of confirmed, hospitalised, and death cases. The model achieved a median MAPE of 61.6% (IQR: 37.6%–79.6%) for confirmed cases and 37.5% (IQR: 18.9%–57.5%) for hospitalisations, and for deaths, the model achieved a median MAPE of 0.65% (IQR: 0.3%–1.9%), indicating substantially higher predictive accuracy for severe outcomes compared to confirmed infections and hospital admissions. Compared to the model without syndromic surveillance data, short-term forecasting performance of the SI_s I_c HRD model improved, reducing MAPE by a median of 15.7 percentage points for confirmed cases, 8.2 percentage points for hospitalisations, and 1.9 percentage points for deaths. Overall, the results demonstrate better predictive stability for severe outcomes, particularly during the early phase of the pandemic.

Conclusions:

The proposed model demonstrated strong predictive performance while remaining simple and computationally efficient. Integrating individual-level data with healthcare helpline syndromic data on testing and severe disease outcomes improves infectious disease surveillance by capturing early signals from symptomatic individuals, allowing us to identify true infections from false alarms, leading to better forecasts of severe outcomes such as hospitalisations. This framework can support public health decision-making in future outbreaks.