Online Symptom Self-Assessment During the COVID-19 Pandemic: An Analysis of Responses from a Canadian COVID-19 Portal


Response Volumes

A total of 34,144 separate responses were collected between April 10 and July 29, 2020. Of these, 1,250 (3.7%) responses were positive for having emergency symptoms and were subsequently referred to the next step in care management. Another 14,340 responses (42.0%) indicated the presence of other non-urgent symptoms potentially related to COVID-19. Another 3,112 responses (9.1%) indicated the presence of high-risk health conditions (eg, heart failure, diabetes, age > 70, autoimmune disease, immunotherapy) without the presence of potential COVID-19 symptoms. 19. Figure 1 shows the number of responses to the Self-Assessment Portal by week. The distribution suggests that the peak of requests on the platform occurred between May 24 and June 28, 2020, with the highest volume occurring on the 3rd June week.

Figure 1

Number of responses to the COVID-19 Self-Assessment Portal over time, Ontario Health West.

Trends in COVID-19 indicators over time

Portal questionnaire responses distinguished between the presence of any one of a cluster of emergency symptoms (e.g., shortness of breath) that would require immediate medical attention and other symptoms that may indicate COVID-19 but do not were not considered to require emergency care. Figure 2 shows the number of portal responses with emergency symptoms present as well as the number of actual cases of COVID-19 reported by the Government of Ontario in postal code (FSA) “N” during the same period. Figure 3 shows the corresponding distributions of any potential COVID-19 symptom (emergency or otherwise) and confirmed cases in the region. In both charts, two y-axis scales are shown to allow comparisons between trends in portal responses and Ontario government-confirmed COVID-19 cases.

Figure 2
Figure 2

Trend in the number of COVID-19 emergency symptoms on the portal compared to actual cases of COVID-19 diagnosed.

picture 3
picture 3

Trend in the number of respondents with symptoms of COVID-19 compared to actual cases of COVID-19 diagnosed.

Geographic Response Patterns

A total of 31,016 (99.6%) responses came from individuals who said they lived in the FSA “N” zone, the main catchment area for which the portal was set up. Figure 4 shows the geographic distribution of response volume and the distribution of symptom counts by FSA area. Since most of the responses came from the “N” region, we limited our analysis to focus on this region. Within the “N” FSA, the N2 region (see online appendix S1 for the table of locations for each sub-region) recorded the highest absolute number of responses with 10,464 responses (Fig. 5) , while the “N7” region had the lowest absolute number with 358 responses over the same period. During the first 5 weeks of data collection, the absolute number of responses from “N6” was the highest, but from the 6th week, the “N2” region recorded the highest numbers per week. During the 11th week, the “N2” FSA recorded an exponential increase in the number of responses (Fig. 6). This may be due to data error or a genuine increase in responses from this region in response to differential COVID-19 outbreak patterns, which will require further investigation to determine the origin.

Figure 4
number 4

Geo-spatial distribution of portal responses and presence of at least COVID-19 related symptoms by F.

Figure 5
number 5

Number of responses to the portal by sub-regions of the “N” postcode area.

Figure 6
number 6

Trend in the number of self-assessments on the portal by postal code.

To calculate comparative response levels between FSAs, we used 2016 Census data from Statistics Canada6 to estimate the number of responses per 1,000 population in each postcode sub-region. FSAs starting with N6 had the highest per capita response, while those starting with N7 both had the lowest per capita response (Fig. 7).

Picture 7
number 7

Portal responses per 1,000 population by postcode subregion.

Symptoms reported by portal respondents

Figure 8 shows trends in reported symptoms over time by symptom type. All symptoms peaked in early May with a decline thereafter; however, a smaller, brief secondary peak was evident in late July. Sore throat was the most common symptom reported by respondents, with 5,880 of 31,016 respondents with non-emergency symptoms reporting having the symptom alone or in combination with other symptoms. The five most reported symptoms (Fig. 9) were sore throat (17.2%), headache (12.9%), fatigue (12.3%), digestive disorders (12.2%). %) and cough (9.1%).

Picture 8
figure 8

Percentage of respondents reporting symptoms related to COVID-19 in FSA ‘N’ zone.

Picture 9
number 9

Frequency of symptoms related to COVID-19 reported on the portal.

In terms of symptom volume, respondents in FSAs N2, N5, and N6 reported the most symptoms, while respondents in FSAs N8 and N9 reported the most average symptom per respondent. However, further analysis showed that the distribution of symptoms within each FSA was broadly similar (Fig. 10).

Picture 10
number 10

Breakdown of the most common types of symptoms reported on the portal by postal code.

Number of symptoms and clusters

There were regional variations in the number of symptoms reported in the Ontario Health West COVID-19 portal. The most common trend was that respondents had recorded no symptoms; however, this is an artifact of two issues. First, if the respondent had emergency symptoms, they were not asked any further questions about other symptoms or their health region. Instead, they were advised to seek immediate medical attention. Second, people who had traveled to areas affected by COVID-19 or who had been in contact with people with COVID-19 were not asked about symptoms. Both subgroups are present in the zero symptom group.

On the other hand, up to 30% of respondents in some sub-regions reported three or more potential symptoms of COVID-19. There are obvious regional variations in the number of symptoms, but care should be taken in interpreting these differences due to the difficulty in differentiating respondents who truly had no symptoms from those who were not. not asked the full set of symptom questions due to the portal’s built-in skip logic.

Using Spearman’s rank sum correlation matrix, we examined patterns of associations between reported symptoms. The results suggest that fatigue tends to be accompanied by chills, headaches, digestive symptoms and sluggishness in children. Not surprisingly, fever and chills were associated with each other, as were congestion and runny nose. Following the implementation of a K-mode machine learning clustering algorithm in the dataset to identify additional clusters, no further clusters were found beyond those emerging from correlational analyzes of the Table 1. It was not possible to perform (supervised) classification machine learning modeling with the data. In part, the performance of machine learning was still limited by the lack of outcome measures or criteria (labels) that could be linked to person-level symptoms.

Table 1 Lead-Lag Correlation Matrix symptoms reported on the portal versus actual COVID-19 cases (per 1,000 population) by sub-region.

Links to primary care

An important function of the COVID-19 Self-Assessment Portal was to connect people with potential symptoms with their primary health care provider, when appropriate. Of the 15,619 responses with symptoms (excluding emergency symptoms), 74.2% (11,589) had a primary care provider to whom they were referred for follow-up. These respondents were able to avoid emergency room visits and were advised to contact the COVID-19 assessment center, as they could be adequately served by primary care based on their response profile. About 17% of respondents with symptoms did not have a primary care provider; however, they were supported by access to an on-call provider.

Relationship between portal symptom reporting and sub-regional outbreaks

Following the initial analysis suggesting similar trends between the number of respondents with symptoms on the portal self-report and the actual number of COVID-19 cases in the region, we conducted further statistical analysis of this. association. We standardized the number of portal symptoms and cases of COVID-19 by calculating each number per 1,000 inhabitants of the respective zip code. To ensure that we covered possible temporal variations of association, we calculated up to 3 weeks lag and primary correlations between portal symptoms and actual outbreaks of COVID-19.

For all sub-regions, the number of people reporting each symptom was overwhelmingly negatively correlated with the number of COVID-19 cases. However, in a few sub-regions, there were significant positive correlations between the number of symptoms reported through the portal and the actual number of COVID-19 cases reported to the Ontario government.

In regions with the first two digits of the postal code, “N4” – “N6”, symptoms inconsistently showed a significant positive correlation with the actual number of COVID-19 cases for the region, with a lag of 2 at 3 weeks. This seems to suggest that in these sub-regions reports of portal symptoms started to increase around 2 or 3 weeks after the increase in actual cases of COVID-19 reported to the Ontario government for each of the sub-regions. This suggests that individuals came to the portal in response to outbreaks that occurred in their sub-region.

Conversely, for sub-regions with the first two digits of the postal code being “N3”, there were positive correlations between some symptoms and actual cases of COVID-19, but with a delay suggesting that people have increasingly started reporting these symptoms on the portal one to two weeks before COVID-19 cases have started to rise in the region. Table 1 shows the symptoms that were found to have positive leading or lagging correlations with COVID-19 cases by subregion. Some sub-regions, however, had no positive association with actual COVID-19 cases.


Comments are closed.