Data collection in pediatric outpatient settings during the SARS-CoV-2 pandemic: motivators and barriers to establishing a research network of primary care providers

Abstract

Background

Data collection in the pediatric primary care settings, conducted by general practitioners and pediatricians, is inherently challenging. However, during the COVID-19 pandemic caused by the SARS-CoV-2 virus, the urgent need for real-time, localized information on the pandemic’s impact on children motivated a network of pediatric practices in the canton of Zurich to undertake systematic data collection and analysis. This initiative aimed to complement established public health networks by providing focused insights from the local pediatric primary care perspective.

Aims of the study

The aim of this study was to establish a research network of pediatric primary care providers (PCP) for data collection in the canton of Zurich, using the pandemic as an opportunity to test the feasibility of such a collaborative system. Secondary aims were to monitor respiratory infections, SARS-CoV-2 test results, and workload impacts among pediatric practices in ther Canton of Zurich during the pandemic.

Methods

Pediatric practices in the canton of Zurich were invited to enter data on respiratory infections, the indications for and results of SARS-CoV-2 tests, and transmission routes into an online database, as well as staff workload and distress per week. Also, initial structural data including the number of staff, hours worked per staff member, and patient volume per week were documented. Structural data were available for all 44 practices, with 28 practices contributing data on respiratory infections and SARS-CoV-2 testing. Following data collection, an online survey was distributed to both participants and non-participants to gather feedback on the study process.

Results

Throughout the data collection period, the weekly number of tests performed varied from 77 to 1066, with positive tests ranging from zero to 65 per week and positivity rates between zero and 0.4. A strong correlation was observed between respiratory infections and the number of tests performed (r = 0.95, p < 0.01) and between the number of tests and the proportion of positive test results in late 2020 and early 2021 (r = 0.492–0.805, p < 0.01). Most infections were attributed to transmission from parents or unknown sources. Structural data indicated wide variations in clinical working hours per week (physicians: median 24 h, range: 8.8–50 h; assistants: median 22.5 h, range: 2.9–63.3) and patient numbers per week (physicians: median 50, range: 3.3–135; assistants: 38.3, range: 7.1–90). No significant correlation was found between the number of SARS-CoV-2 tests and staff stress levels. Feedback on the data collection indicated that participants found data entry manageable, highly valued the regular interim feedbacks on results, and expressed strong interest in participating in similar future studies.

Conclusion

The study demonstrated the feasibility of regional data collection within a cantonal network, yielding up-to-date epidemiologic insights during the pandemic. Participants were highly motivated, with regular feedback on interim results cited as a key incentive for continued engagement.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12875-025-02934-7.

Introduction

In Switzerland, primary pediatric care is predominantly provided by pediatricians in private practice and general practitioners [1]. Over 1,300 pediatricians, accounting for 61% of the pediatric workforce, are involved in primary care. However, despite broad recognition of the importance of outpatient research by several Swiss academic and primary care organizations, such as the Kollegium für Hausarztmedizin (KHM), SwissPedNet, Kinderärzte Schweiz (KIS), and Swiss Pediatrics, data from the pediatric outpatient care sector remain underutilized for research purposes. Existing initiatives have included only limited contributions from pediatric outpatient settings [2, 3]. The primary aim of this study was to establish a research network of primary care pediatricians (PCPs) in the canton of Zurich, to gain practical experience with collaborative data collection, and to test the feasibility of such a network. These efforts were intended to lay the groundwork for expanding and institutionalizing a nationwide PCP research network. In building this network, the study also aimed to inform future research initiatives through feedback gathered from participating practices on the data collection process.

During the global SARS-CoV-2 pandemic, PCPs played a critical role in patient care and SARS-CoV-2 testing for children. However, there remains a significant lack of centrally collected pediatric outpatient data available for research purposes in Switzerland [4]. Consequently, data on the prevalence of SARS-CoV-2 infection were limited to aggregate levels (i.e., cantonal or national), without specific information such as rates of positive test results in children, modes of transmission, or reasons for testing [5]. PCPs in the canton of Zurich were therefore highly motivated to participate in a network aimed at collecting and analyzing timely, region-specific data on SARS-CoV-2 prevalence in children, including positivity rates, modes of transmission, and reasons for testing. While some aggregated data were available at the cantonal level, these datasets lacked the granularity needed for informed, local-level decision-making [5]. PCPs across Switzerland expressed a need for better information about the pandemic’s dynamics within their immediate environment, since access to such real-time information could have allowed practitioners to respond effectively to emerging challenges, anticipate potential burdens, and ensure continuity of care. However, official updates from government agencies often did not meet these needs. Moreover, there is a dearth of data on the operational aspects of pediatric practices, such as the number of physicians or staff per practice, average hours worked per week, and the number of patients seen per week [6]. Finally, the extent to which the pandemic contributed to increased workload and distress among PCPs and their teams remained unclear. To address this, we analyzed whether the number of SARS-CoV-2 tests performed (both antigen and PCR) correlated with the number of respiratory infections recorded in participating practices. We also assessed the weekly number of tests performed per practice, the proportion of positive tests, and the most frequently reported routes of transmission. Furthermore, we examined whether tesing volume or infection rates were associated with the subjective workload reported by practices.

The pandemic also presented an opportunity to establish a local PCP research network capable of generating real-time, region-specific data. By focusing on pandemic-related outcomes, including infection trends, transmission patterns, test results, and workload dynamics, this study aimed to demonstrate the practical value and feasibility of such a network as a foundation for broader, long-term research infrastructure.

Methods

We conducted a longitudinal observational/epidemiological surveillance study among pediatric practices in the canton of Zurich during the SARS-CoV-2 pandemic, and a consecutive evaluation by an online-survey five months later. The primary data collection period spanned from week 43 of 2020 to week 33 in 2021 (October 19, 2020, to August 10, 2021). Data were collected using a centralized REDCap database and analyzed bi-weekly to track respiratory infections, SARS-CoV-2 tests performed, positive test results and the workload of participating practices [7].

Recruitment and participants

Email invitations were sent to all pediatric practices in the canton of Zurich using the mailing list of the Association of Zurich Pediatricians (VZK), which included approximately 300 members, some working in group practices, and some alone in private practices in 2021.

Data collection

Data collected included the number of respiratory infections (defined as an acute infection of the respiratory system, including the upper or lower respiratory tract, caused by various pathogens and resulting in symptoms such as cough, fever, and difficulty breathing), the number and type of SARS-CoV-2 tests (PCR or antigen), test indications, patient age, and– for positive cases– symptoms, patient sex, and transmission routes reported by families. Additional data included structural data such as the practice postcode, weekly clinical working hours of physicians and assistants/nurses (clinical and administrative work), number of patients seen per week, subjective workload (“how was your workload during the last week”, self-rated on a 4-class Likert-type scale, ranging from zero to very high, which was created for this study), and the number of staff absent due to COVID-19. Data were self-reported and entered weekly into the REDcap database [7]. Participants received weekly updates summarizing interim results to provide a real-time overview of pandemic dynamics in their region as an incentive for participation.

Survey for participant feedback

An online survey was administered in April 2022, five months after data collection was concluded, to gather feedback from participants and non-participants. The survey, which took approximately ten minutes to complete, collected information on general feedback, motivations for participation, and interest in future studies. It also addressed specific aspects such as technical issues during setup, the usability of REDCap instructions [7], the data entry process, and additional time requirements for participation. It was distributed via the same address list as the invitation to the initial data collection, to which the accompanying mail also referred.

Efforts to minimize bias

To reduce selection bias, all pediatric practices affiliated with the regional pediatric organization were invited to participate. However, since no official register of pediatric practices is publicly available in the canton (e.g., by the health department or statistical office), it was not possible to reliably assess the representativeness of the sample. We estimate that the actual number of individual practices is approximately one-third lower than the number of pediatricians, as many work in group practices. Nonetheless, the use of self-reported data introduced the potential for reporting bias, particularly during periods of staff absence, holidays, or increased workload, when data entry may have been deprioritized.

Statistical analysis

Descriptive statistics were utilized to summarize the demographic and clinical characteristics of the practices. Continuous variables were reported as means with standard deviation or median with ranges, depending on the normality of the distribution (assessed via the Shapiro-Wilk test). Categorical variables were presented as frequencies and percentages. Key variables were visualized using histograms and boxplots. No imputation was performed for missing data, and outliers were identified using z-scores. Correlations between respiratory infections, the number of tests performed, and the number of positive tests were analyzed, using pearson-correaltions, with a p-value of < 0.05 considered statistical significant. All analyses were conducted using R © software (version 4.1.2; 2021-11-01) within the R-Studio © environment [8].

The study was conducted in accordance with the Declaration of Helsinki. The cantonal ethics committee of Zurich reviewed and approved the study protocol. A waiver of consent was granted as only anonymized data were collected (Req-2020-01256 and Req-2020-01256). The protocol was not formally registered but can be made available upon request. Due to data protection regulations, study data can only be shared in aggregated form upon reasonable request.

Results

A total of 55 practices across the canton responded to the email invitation and agreed to participate. The geographical distribution of the 55 participating practices shows a concentration in urban and peri-urban areas of the canton.The city of Zurich was most strongly represented with 13 participating practices. Other relevant locations were Horgen and Meilen with 5 practices and Winterthur with 4 practices. The following municipalities also participated in the study, each with between one and three practices Affoltern am Albis (1), Birmensdorf (1), Bülach (1), Dietikon (3), Dielsdorf (1), Dübendorf (1), Hinwil (1), Männedorf (1), Ottenbach (1), Richterswil (1), Uster (3), Wädenswil (1), Zollikon (1). 11 of the 55 participating practices did not state an exact location. Some practicioners had already initiated local data collection efforts independently and expressed strong motivation to join a coordinated network. Initial challenges included technical implementation issues in certain practices and the time required to collect and enter data. As an incentive, all PCPs received bi-weekly information letters containing responses to frequently asked COVID-related questions, as well as interim results from the network’s data collection. Practices were excluded from the final analysis if they submitted data for fewer than 15 weeks.

COVID-related data were collected from 28 practices, while structural data were obtained from 44 practices out of the 55 participating practices, with one practice contributing only COVID-related data and no structural data.

A total of 18,013 patients with respiratory infections were reported, and 11,456 SARS-CoV-2 tests were conducted, of which 433 were positive. This resulted in an overall positivity rate of 3.78%, with a peak observed during calendar weeks 50 and 51. The weekly number of tests performed varied widely, ranging from 77 to 1066, while the number of positive tests ranged from 0 to 65 per week.

A strong correlation was identified between the number of respiratory infections and the number of tests performed (r = 0.95, p < 0.01), with parallel fluctuation in these variables observed over the study period (Fig. 1).

Fig. 1.

Number of respiratory infections, tests performed, and positive tests. Number of respiratory infections A, numbers of tests performed B, and number of positive tests C by calendar week. Total number of tests is subdivided into testing by PCR (dark blue) and antigen (light blue). Total number of positive test results is subdivided by PCR (dark red) and antigen (orange). The number of practices for which data was available in each week is indicated as the grey bars

A correlation was also observed between the number of tests performed and the number of positive test results, as shown in Fig. 1. At the end of 2020, there was a strong positive correlation between the number of tests performed and the number of positive results (r = 0.805, p < 0.01, based on 11,456 tests). However, this correlation was notably weaker during the first three months of 2021 (r = 0.492, p < 0.01). A comparison across age groups revealed that younger children underwent significantly more tests than older children and adolescents. The mean number of tests conducted for children under 12 years was 43.03 (SD = 75.62), compared to 13.64 (SD = 724.7) for older children and adolescents.

Most infections were reported as being acquired from parents or an unknown source (Fig. 2). Less frequently, infections were attributed to siblings or other individuals. School exhibited an intermediate level of transmission, with uncertainty surrounding whether the source of infection was a classmate or a teacher.

Fig. 2.

Modes of infection and possible combinations. The X-axis represents different modes of infection and possible combinations, while the Y-axis shows the number of infections for each mode. Horizontal bars on the left show the number of cases for each infection mode. Vertical bars represent the individual combinations, and horizontal bars represent the individual options combined

Figure 3 illustrates the wide variation in workload levels among participating practices. While a few practices, particularly one, reported consistently high workloads, most reported moderate to lower workload levels. Periods of increased testing generally coincided with increased workload levels. For example, between weeks 10 and 20 of 2021, there was a notable spike in testing, with over 1000 tests conducted in week 13. During this period, a slight increase in workload was observed across practices, with three practices reporting significant high workload leves. However, when analyzed across the entire study period, no statistically significant correlation was found between the number of tests performed and stress levels.

Fig. 3.

Number of SARS-CoV-2 tests performed and workload of practices. A SARS-CoV-2 tests (PCR- and Antigen-tests) performed in all practices as numbers per calendar week in 2020) and 2021), B subjective workload (self-rating on a 4-category scale ranging from none to strong) of each participating practice during the same weeks. Tests performed are summed up for all practices, whereas the subjective workload is displayed individually for each practice

Table 1 provides an overview of the structural data of the 44 participating practices, including the number of physicians and assistants per practice, weekly hours worked by each employee, and the number of patients seen. The data reveal a significant variation in all variables across practices.

Table 1.

Structural data of participating practices (n = 44). Showing the median, maximum, minimum and interquartile range for each professional group in number of empoyees, working hours per week and patients per employee

	Median	Maximum	Minimum	Interquartile range
Number of physicians per practice	3	16	1	2
Number of assistants per practice	4	24	1	3
Number of assistants per physician	1.5	6	0.33	1
Weekly working hours per physician (clinical)	24	50	8.75	19.35
Weekly working hours per assistant (clinical and administrative)	22.5	63.33	2.84	9.19
Patients per physician per week	50	135	3.33	28.75
Number of patients per assistant per week	38.33	90	7.14	21.25

A total of 28 practices contributed SARS-CoV-2 data, of which 20 also participated in the survey. In addition, 12 practices that did not contribute to the initial data collection joined the survey, bringing the total number of survey participants to 32 pediatric practices (Table 2). Most practices reported no difficulties with the installation and data entry processes, with only a few suggesting areas for improvement in future studies.

Table 2.

Results of the survey evaluating the data collection

	Yes	No	Not answered
Participation	20	12
Setting up and Registration
Were the e-mail instructions helpful?	17	2	13
Did the installation of the account work well?	15	4	13
Were there any problems with the application?	3	17	12
Data entering
Were there any problems with the data entering?	4	16	12
Did you regularly get feedback with interim results?	18	1	13
Did you have written down data, not entering them to Redcap?	7	9	16
Time effort
Was the time effort fine?	15	5	12
Did technical complications rise the time effort?	4	15	13
Was it an additional organisational effort to take part in the study?	8	12	12
Suggestions for future studies
Would you have wished for a better instruction to the program?	4	16	13
Would you have wished for a point of contact for problems?	4	14	14
Would you have wished for regularly feedback?	8	10	14
Would you have wished for regularly request about problems or difficulties?	3	15	14
Willingness to take part in further studies	9	0	23

Total number of participating practices = 32

The primary motivation for participation was the need for timely and region-specific information, cited by 14 practices (70%). Some respondents (n = 3, 15%) expressed an interest in gaining a direct overview of the pandemic’s impact on pediatric primary care, while only two practices (10%) participated to compare the pandemic situation in their region with that of other regions.

Participants proposed several potential areas for future research, including practical initiatives such as assessing cancer treatment outcomes in pediatric patients and investigating the administrative workload in primary care settings. They also highlighted the importance of assessing psychosocial concerns, such as educational challenges and child protection issues. Specific topics, such as antibiotic resistance and diseases related to bacterial infections, were also suggested. The most commonly cited reason for non-participation in the study was a lack of time.

Discussion

Our study aimed to establish a research network of primary care pediatricians to enable systematic, prospective data collection within a regional cluster. We were able to demonstrate the feasibility of organizing such a network of PCPs, which successfully collected valuable, practice-based data over the course of several months. Our study successfully collected data on SARS-CoV-2 infections within a regional research network of pediatric practices in the canton of Zurich during the pandemic. This initiative helped close the gap in granular pediatric SARS-CoV-2 data by providing detailed insights into infection rates, modes of transmission, and the number of SARS-CoV-2 tests performed. Notably, the prevalence of respiratory infections did not significantly impact the subjective workload of primary care pediatricians. Furthermore, we demonstrated the feasibility of collecting outpatient data, including structural information about participating practices. The strong engagement of PCPs reflected their motivation to contribute to a detailed regional understanding of the pandemic and their interest in accessing timely information.

The variability in the number of SARS-CoV-2 tests performed and positivity rates observed in our data aligns with cantonal statistics [9] but provides a more nuanced perspective when compared with respiratory infections and transmission modes. This granularity was highly valued by participants and highlights the potential for conducting similar research projects in the future through regional PCP networks. While the correlation between respiratory infections and testing volume was strong, as expected, no statistically significant correlation was observed with the subjective workload of PCPs [9]. Testing trends remained stable during periods of strict public health measures [10] but increased following the relaxation of restrictions [11] and the start of summer vacations [12]. These trends parallel cantonal data [13] and reflect the introduction and subsequent rise in antigen testing after approval by the Federal Office of Public Health (FOPH) on November 2 [14]. In our study, the most frequent transmission mode was via parents or unclear, less frequently via school or siblings.

The observation that fewer tests were performed in children older than 12 years aligns with a higher incidence of respiratory infections in younger children [15]. However, there is limited consensus in the literature regarding SARS-CoV-2 incidence and testing rates among adolescents aged 12 to 17 years compared with younger children [16, 17]. The introduction of the COVID-19 vaccine in Switzerland in December 2020, initially recommended for adolescents aged 16 years and older and adults [18], may have also influenced testing patterns, as most infections in children were attributed to transmission from adults, consistent with existing research [19].

The structural data collected revealed significant heterogeneity in pediatric practices, particularly in the number of physicians, staff, weekly working hours, and patients seen per week. Variations in working hours compared to those reported in the literature [6, 20] highlight the substantial administrative burden in primary care settings. Administrative tasks, which account for approximately 4.7% of total working hours for general practitioners [6], likely contribute to these discrepancies. Despite this variability, the structural characteristics observed appear suitable for future comparative studies of private practices.

We observed strong engagement from the pediatric community, with 55 practices agreeing to participate out of approximately 300 pediatricians in the region– many of whom work in group practices. In 2017, for instance, there were 115 pediatric practices serving the canton) [21]. This high response rate reflects the high intrinsic motivation among PCPs and the perceived relevance of the study topic. However, only 28 practices completed the full data collection, underscoring the practival challenges such a network faces, especially when participation requires additional time and effort without financial compensation. This highlights the need to consider resource allocation and sustainability when designing future research initiatives in primary care settings. The distribution of participating practices in the canton of Zurich demonstrates a marked concentration in urban and peri-urban regions. A total of 42 General Practice (GP) practices from the canton participated in the study, with the majority of these practices located in urban areas. Of particular note is the city of Zurich, where 12 practices were represented, accounting for around 22% of the total number of participating practices. The high rate of participation observed in the city of Zurich may be indicative of a tendency among urban regions to exhibit a greater propensity to engage in such research endeavours. This phenomenon could be attributed to enhanced accessibility of resources, specialists, and infrastructural facilities, which are often found in urban regions.

Survey results demonstrated that PCPs’ motivation to obtain comprehensive, localized information about the epidemic outweighed the time and effort required for data collection. Participants underscored the need for accessible and user-friendly data collection methods. However, our study also revealed the critical need for financial resources to support coordination and data management, which limited the duration of data collection. Established pediatric outpatient research networks in other countries, such as PROS and PECARN [22], serve as valuable models for integrating routine data into scientific research, particularly with adequate funding. In Switzerland, initial steps in this direction include the SwissPedHealth project, “Preparing PERsonalizEd PediatRic PrimaRy care (PREPP)” [23], and the “Child and Adolescent Health - Minimal Set of Indicators for Switzerland” project initiated by the Federal Office of Public Health (FOPH) [24].

Strengths and limitations

Our study demonstrated the feasibility of establishing a regional PCP network for systematic data collection in the outpatient setting. The persistent and regular data entry by several participating practices reflected the high motivation of PCPs in Zurich to contribute to a more detailed and up-to-date epidemiological understanding of the COVID-19 pandemic in children. However, the sample size for the study was limited due to time constraints and the absence of financial compensation for the additional workload, and we are not able to comment on potential selection biases, or if there were differences in characteristics between the participating and non-participating centers (e.g., location) since there is no objective statistics on number or local distribution of pediatric offices available in Switzerland. However, since group practices are a frequent phenomenon in the Canton, the number of offices must be assumed to be less than the number of physicians organized in the professional association of pediatricians. The regional focus of the study may restrict the generalizability of findings to other parts of Switzerland. Nonetheless, comparable initiatives exist in other regions, and we are confident that the approach presented here could be expanded into a multi-site, and ultimately national, network. A further limitation was the high proportion of unanswered items in the follow-up questionnaire, which limits the intepretability of the feedback. In addition, variability in reported working hours and the lack of distinction between full-time and part-time positions made it difficult to compare weekly patient volumes across practices. Future studies should address these limiations and investigate whether unanswered survey questions reflect a lack of interest, time constraints, or other underlying barriers.

Conclusions

Our study successfully initiated a regional PCP research network and demonstrated the feasibility of systematic, prospective data collection during the COVID-19 pandemic. The network generated timely, practice-based regional insights into the pediatric impact of the pandemic, complementing existing cantonal and federal data. Importantly, it showed that structured data collection is achievable in pediatric outpatient settings. Feedback from participants emphasized the importance of addressing relevant research questions, minimizing administrative burden, and providing rapid, meaningful feedback as key factors to sustaining PCP engagement in future research initivatives.

Authors’ contributions

A.L has made the data analysis and prepared the figures. A-L, M.vR, J.T wrote the manuscript text. The manuscript was revised by M.S, C.B, R.vdH, A.BS, M.S.

Funding

Kinderärzte Schweiz (Swiss professional association of pediatrics in private practice) supported the study with a small amount to cover ethics.

Data availability

Due to data protection regulations, study data can only be shared in aggregated form upon reasonable request.

Declarations

Ethics approval and consent to participate

The study was conducted in accordance with the Declaration of Helsinki. The cantonal ethics committee of Zurich reviewed and approved the study protocol. A waiver of consent was granted as only anonymized data were collected (Req-2020-01256 and Req-2020-01256).

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.