Realtime Supervision and Quality Assessment of Statewide Community based Sero-Surveillance for Anti-Sars-Cov-2 IgG Antibodies

Introduction

In sero-surveillance survey done in Haryana, smartphone and a survey app was used in addition to traditional paper-based format to report data for the sero-surveillance of COVID-19 in all 22 districts [1]. We used the GPS app to report observations from our on-site supervisory visits, done for quality assurance, on a real-time basis. Here we present a model of supportive supervision for quality control of seroprevalence survey that may be of use to plan quality assurance for other such surveys at the state and the national level.

Methodology

Seroprevalence survey for anti-SARS-CoV-2 IgG antibodies in all 22 districts of Haryana was planned from 17 August to 19 August 2020. All the districts were to complete the survey within three days. Using digital mode, survey tools were finalized in consultation with Haryana state team. In order to maintain the quality of sero-surveillance, we planned to monitor and supervise the entire surveillance cycle. The complete cycle of sero-surveillance consisted of:

• A). Orientation training of state surveillance team
• B). Supervisory Plan and Sampling for Quality Assurance
• C). Visit to sites and testing laboratories
• D). Real time monitoring
• E). Review Meetings with state core team

Results

We visited 13 clusters in 8 districts, and observed 28 teams in the community and 4 district laboratories. Most of the clusters (92%), and all laboratories were graded satisfactory (score 8 or above).

Most survey teams conducted the survey within the ambitious time span of 3 days. Practically entire state teams were in the field despite torrential rains throughout the state. Teams did follow the survey methodology with little deviations at some places explained later. As per the guidelines, survey teams had divided each cluster into quadrants and samples were collected from each quadrant. There was a doctor in each team. Internal supervisory teams of the districts and state were also in the field. Most of the android phone users were able to use the survey app. Medical officers took the informed consent to a large extent as per the trainings given to them. Household IDs were written correctly to a large extent on the survey forms as well on the blood sample vials. Blood samples were collected, labelled and dispatched to the laboratories on the same day. For blood sample collection, at most clusters a good protocol for asepsis was followed. At many places biomedical waste management guidelines were also followed. (Table 2)

Three of the 4 testing laboratories were fully operational and prepared for the incoming samples.

Some gaps in the conduct of the survey also came to our notice. There was intra and inter district variation in the randomization technique to select the study sample. One cluster team carried out randomization by sampling every third house, while in another cluster consecutive door-to-door visits were done to complete the sampling. In some clusters, sampling was done by calling the families at a common convenient place due to many operational difficulties including heavy rains. The selection of subject from each household was to be done on the basis of a specified grid table. The teams in some clusters were having trouble in interpreting the grid table. Survey app did not have the option to use the grid table for sampling. In certain cases, the MOs were proactively guiding the team to correctly read the grid table, in the others we directed the teams in this regard. Due to operational difficulties with the online app and due to network issues, some MOs were finding it difficult to upload the form on the portal.

There were some instances of spilling of blood sample, and non-wearing of gloves by LTs. In one cluster the LT was moving around with an open used syringe. In certain clusters only yellow-biomedical waste management bags were available. In two clusters hub-cutter was not there. These teams were found practicing re-capping of the needles. At one cluster, team was struggling to draw the blood sample and was reluctant to follow standard operating procedures as they seemed to be untrained. The lab-technician from our team then demonstrated blood sample collection on-site. One of the testing laboratories, was not equipped with the requisite ELISA kits.

Discussion

The present study demonstrates importance of real-time supervision and quality assurance in a massive and rapid survey of national importance. The serosurvey was planned for all 22 districts to be completed in 3 days-time. It was not possible to depute independent teams for quality assurance in all 22 districts. Therefore, multipronged approach was followed by having mixed model of supervision with both internal and external agencies. It helped to build the inhouse capacity and also gave credibility to the survey findings.

A mixed type of supervision along with real-time monitoring, and use of geo-coding to locate the house were the main features of the study. Our study showed that quality of sero-surveillance was satisfactory in most of the clusters (92%) and in the laboratories. This study also showed that quality assurance helps to take last minute corrective actions and build the capacity if required. Those who are found performing well also get a moral boost when they are observed and appreciated.

Though the quality of sero-surveillance was found more than satisfactory, authors also noted certain gaps. At some places untrained staff was deployed. This survey was done during active Covid-19 pandemic when everyday some health staff was getting infected or getting quarantined. Thus, some trained staff due to infection had to take emergency leave and had to be replaced by the untrained staff at some places, as the mass survey was time bound and had to be completed within short period of 3 days. Others have also reported the link between the training of staff and survey quality [2, 3]. Districts had ensured deployment of trained staff along with the untrained and supervisors were also vigilant, still the impact on quality is inevitable. Even trained staff can commit errors as was found in this survey where some trained staff was found doing error in sample collection. Independent quality assurance helped to take cognizance of this and note that magnitude of these errors was miniscule and even these errors were addressed during quality assurance visits.

Some gaps in deviation from the protocol were also circumstantial. In some communities, during the rains, it was difficult to move from house to house. Thus, field teams tried to find a sheltered place to sit in the sampled area and called the household members one by one to the place. They maintained privacy. The families needed to move just to the neighborhood. Thus, these types of novel adjustments did not influence the sanctity of the sampling method. Rather, the quality assurance visits brought to the light the passion of the field team, who were not deterred by the heavy incessant rains, and who also tackled the community resistance. Some families were reluctant to allow any health team member to enter their house due to the covid fear. Realtime monitoring helped us to track this and the approvals of such deviations were granted in real time. These gaps cannot be considered as absolute gaps but are the ground zero adjustments in the given context.

Most of health care providers were using universal precautions but disposal of bio-medical waste was found inadequate/ inappropriate in the present study as well. This necessitates a training programme for waste management in future [2]. We observed a lack of trained laboratory technicians and ELISA kits. These findings were similar to those observed by Panda B during supportive supervision of immunization session Odisha, India [4]. Authors suggest that supportive supervision along with handling the systemic issue, such as testing kits, trained health staff and timely indenting can improve the quality of COVID-19 sero-survey.

There were operational difficulties in the use of surveyman app, but with the involvement of app managers, most of the difficulties were resolved. Yet, the app could not provide access to each district supervisor and state supervisor to locate the survey teams and have independent control of the survey teams. Only Principal Investigator had this access. Due to rapid launch of the survey, supervisors, except the PGIMER team, were not well trained to use the app and thus they followed traditional method of supervision to a large extent. The app didn’t work for ‘ios’ users. Quality assurance helped to bring such gaps to the notice of Principal Investigator who resolved most of the gaps in realtime by personal interaction with the surveyors and app manager.

Similar to the present study, Biemba G et al in their study done in two districts of Zambia observed that mobile app can be effectively used for the supervision of community health workers and delivering quality services [5]. A study was conducted by Henery JV et al. using Whatsapp group chats and photos of 1830 posts and 41 participants for six months. The authors demonstrated that use of mobile app can be an effective channel of communication for supervision and support at community level [6]. The added advantage of using global positioning system (GPS) to locate house for real-time monitoring and for taking action was used in present study similar to Noor AM et al. in Kenya and Sisay G et al. in Nigeria for finding distance travelled by patients for getting services from government health services [7, 8].

Thus, one of the greatest advantage real time monitoring over traditional supervision was visibility of the status of each survey team with immediate action on any issue, if found. Also, incomplete or incorrect survey/forms were detected in time before the survey lost its desired qualities. Results have shown that the use of such a system can provide a solution that tracks the teams and assesses the data quality in real time, and can provide immediate solutions to the teams which is especially important for the large population based studies.

Conclusion

App-based, real-time monitoring, along with GPS was successfully used for monitoring the quality of community based sero-surveillance for anti-SARS-CoV-2 IgG antibodies. Countries can use our findings to measure and improve the quality of sero-surveillance as well as for the large population based studies.

References

1. Aggarwal AK, Sharma A, Kumar D. Use of Mobile App for Real Time Data Entry and Monitoring: A Case Study of Anti-SARS-CoV-2 IgG Antibody Sero-Surveillance in all Districts of Haryana, India. International Journal of Health Systems and Implementation Research. 2020;4(2):63-9. [Google Scholar]
2. Kant S, Goswami K, Rai S, Dar L, Misra P. Quality assurance of annual HIV sentinel surveillance 2006: Experience of regional institute, central zone. Indian J Public Health. 2007;51(1):28-32. [Google Scholar]
3. Momanyi GO, Adoyo MA, Mwangi EM, Mokua DO. Value of training on motivation among health workers in Narok County, Kenya. Pan Afr Med J. 2016;23(1). [doi:10.11604/pamj.2016.23.261.8414]
4. Panda B, Pati S, Nallala S, Chauhan AS, Anasuya A, Som M, et al. How supportive supervision influences immunization session site practices: a quasi-experimental study in Odisha, India. Global health action. 2015;8(1):257-72. [doi:10.3402/gha.v8.25772]
5. Biemba G, Chiluba B, Yeboah-Antwi K, Silavwe V, Lunze K, Mwale RK, et al. A mobile-based community health management information system for community health workers and their supervisors in 2 districts of Zambia. Glob Health Sci Pract. 2017;5(3):486-94. [doi:10.9745/ghsp-d-16-00275]
6. Henry JV, Winters N, Lakati A, Oliver M, Geniets A, Mbae SM, et al. Enhancing the supervision of community health workers with WhatsApp mobile messaging: qualitative findings from 2 low-resource settings in Kenya. Glob Health Sci Pract. 2016;4(2):311-25. [doi:10.9745/ghsp-d-15-00386]
7. Tegegne SG, Shuaib F, Braka F, Mkanda P, Erbeto TB, Aregay A, et al. The role of supportive supervision using mobile technology in monitoring and guiding program performance: a case study in Nigeria, 2015–2016. BMC public health. 2018;18(4):75-81. [Google Scholar]
8. Noor AM, Amin AA, Gething PW, Atkinson PM, Hay SI, Snow RW. Modelling distances travelled to government health services in Kenya. Tropical Med Int Health. 2006;11(2):188-96. [doi:10.1111/j.1365-3156.2005.01555.x]