Home About us Editorial board Ahead of print Current issue Archives Submit article Instructions Subscribe Contacts Login 
  • Users Online: 455
  • Home
  • Print this page
  • Email this page

 Table of Contents  
Year : 2020  |  Volume : 27  |  Issue : 4  |  Page : 280-284

Predictive ability of symptomatology in COVID-19 during Active case search in Lagos State, Nigeria

1 Lagos State Primary Healthcare Board, Lagos, Nigeria
2 Lagos State University College of Medicine, Lagos, Nigeria
3 Nigerian Centre for Disease Control, Lagos, Nigeria
4 Lagos State Ministry of Health, Lagos, Nigeria
5 Lagos State Health Management Agency, Lagos, Nigeria
6 Lagos State University Teaching Hospital, Lagos, Nigeria
7 Mainland Hospital, Yaba Lagos State, Lagos, Nigeria
8 Lagos University Teaching Hospital, Lagos, Nigeria
9 World Health Organisation, Nigerian Office, Lagos, Nigeria
10 College of Medicine, University of Lagos, Lagos, Nigeria
11 African Field Epidemiology Network, Lagos, Nigeria

Date of Submission25-Jul-2020
Date of Decision01-Aug-2020
Date of Acceptance16-Sep-2020
Date of Web Publication04-Nov-2020

Correspondence Address:
Prof. Akin Osibogun
Department of Community Health and Primary Care, College of Medicine, University of Lagos, Idi-Araba
Login to access the Email id

Source of Support: None, Conflict of Interest: None

DOI: 10.4103/npmj.npmj_237_20

Rights and Permissions

Background: In April 2020, a community-based active case search surveillance system of coronavirus disease 2019 (COVID-19) was developed by the emergency outbreak committee in Lagos State. This followed the evidence of community transmission of coronavirus disease in the twenty Local Government Areas in Lagos State. This study assessed the value of respiratory and other symptoms in predicting positive SARS-CoV-2 using reverse transcription-polymerase chain reaction (RT-PCR). It is hoped that if symptoms are predictive, they can be used in screening before testing. Methods: Communities were included based on the alerts from community members through the rumour alert system set up by the state. All members of the households of the communities from where the alert came were eligible. Household members who declined to participate were excluded from the study. A standardised interviewer-administered electronic investigation form was used to collect sociodemographic information, clinical details and history for each possible case. Data was analysed to see the extent of agreement or correlation between reported symptoms and the results of PCR testing for SARS-COV-2. Results: A total of 12,739 persons were interviewed. The most common symptoms were fever, general weakness, cough and difficulty in breathing. Different symptoms recorded different levels of sensitivity as follows: fever, 28.9%; cough, 21.7%; general body weakness, 10.9%; and sore throat, 10.9%. Sensitivity and specificity for fever, the most common symptom, were 28.3% and 50.2%, respectively, while similar parameters for general body weakness, the next most common symptom, were 10.9% and 73.2%, respectively. Conclusion: From these findings, the predictive ability of symptoms for COVID-19 diagnosis was extremely weak. It is unlikely that symptoms alone will suffice to predict COVID-19 in a patient. An additional measure, such as confirmatory test by RT-PCR testing, is necessary to confirm the disease.

Keywords: COVID-19, diagnostic accuracy, SARS-CoV-2, symptoms

How to cite this article:
Onasanya O, Adebayo B, Okunromade L, Abayomi A, Idris J, Adesina A, Aina O, Zamba E, Erinosho O, Bowale B, Opawoye F, Ramadan P, Yenyi S, Omilabu S, Balogun S, Osibogun A. Predictive ability of symptomatology in COVID-19 during Active case search in Lagos State, Nigeria. Niger Postgrad Med J 2020;27:280-4

How to cite this URL:
Onasanya O, Adebayo B, Okunromade L, Abayomi A, Idris J, Adesina A, Aina O, Zamba E, Erinosho O, Bowale B, Opawoye F, Ramadan P, Yenyi S, Omilabu S, Balogun S, Osibogun A. Predictive ability of symptomatology in COVID-19 during Active case search in Lagos State, Nigeria. Niger Postgrad Med J [serial online] 2020 [cited 2021 Apr 11];27:280-4. Available from: https://www.npmj.org/text.asp?2020/27/4/280/299910

  Introduction Top

The coronavirus disease 2019 (COVID-19) poses a major threat to health, policy, lifestyle and the economy globally.[1] According to the Johns Hopkins Center for Systems Science and Engineering, as of 23 August 2020, there were over 23,293,558 confirmed cases and 806,294 deaths globally.[2] On the same date, Nigeria had recorded 51,905 confirmed cases of COVID-19 with 997 deaths, while, in Lagos State, the epicentre of the pandemic in Nigeria had 17,764 confirmed cases and 277 deaths.[3] The first case of COVID-19 managed in Lagos State was an Italian who flew into Lagos through Istanbul on 25 February 2020 and became ill while in Ogun State on a business trip but brought back to Lagos by 27 February when his sample for SARS-COV-2 testing came out positive.

Human coronavirus transmission occurs through droplets and indirect or direct contact.[1],[4] SARS-COV-2, similar to other respiratory viruses, has the highest transmissibility when patients are symptomatic, as viral load peaks shortly after the onset of illness.[4] Early signs and symptoms of COVID-19 disease include fever, cough, shortness of breath and myalgia.[4],[5] Other symptoms include nasal congestion, chills, sore throat, anosmia and diarrhoea; current evidence suggests that symptoms occur within 2–14 days after exposure.[4],[5] About 80% of those affected present with mild forms of the disease and 15% with moderate-to-severe disease.[5] However, older patients and those with comorbidities or immunocompromised status have a higher risk of developing severe disease.[6],[7],[8]

To break the transmission of the disease in Lagos State, there was an urgency to identify cases for isolation and understand the severity and proportion of illnesses and monitor trends and burden of disease in other to flatten the epidemic curve by active case ascertainment using contact tracing.

It was hoped that if the symptoms were highly predictive of the SARS-COV-2 status in COVID-19 contacts, the use of symptomatology algorithms could be a welcome solution to the challenges currently experienced with getting enough testing (extraction) kits.

  Methods Top

This study used data collected during the active case search effort of the emergency response to the outbreak in Lagos State and is, therefore, a cross-sectional descriptive study.

Lagos State is located in the South-western geopolitical zone of Nigeria.[9],[10] It has the smallest area of Nigeria's 36 states. Its estimated population according to the 2006 National Census was put at 9,013,534.[10] Ogun State shares its border with the north and east side of Lagos State and the Republic of Benin on the west side of Lagos State.[9],[10] The state is divided into twenty Local Government Areas (LGAs) and 37 Local Council Development Areas. Lagos State has 376 wards.[9]


A person was classified as a probable case of COVID-19 infection if he or she was a household, family or health-care contact of a person with a confirmed case and if pneumonia developed without another confirmed cause and either laboratory testing was not carried out, or the person died having presented with fever, cough and difficulty in breathing.

A person was classified as a suspected case of COVID-19 infection if they have an acute respiratory infection with at least one of the following symptoms: cough, fever or difficulty in breathing.

A person was classified as a confirmed case of COVID-19 infection if there was laboratory evidence of SARS-COV-2 with or without symptoms.

The date of onset was defined amongst febrile patients as the 1st day of fever that persisted for more than 48 h and amongst a febrile respiratory patient as the 1st day of new cough or shortness of breath.

A person was considered to have been exposed if he or she had had any face-to-face contact with an asymptomatic patient who had a confirmed or probable case.

Close contact was defined as anyone who had prolonged contact within 2 m distance to the case.


The active case search was carried out for 4 weeks to identify those with symptoms and refer them for polymerase chain reaction (PCR) testing. Health workers and community informants were selected within the wards. They were trained on infection prevention and control and the use of Open Data Kit (ODK). ODK is an open-source phone-based platform for collecting and aggregating data.[11],[12] This is illustrated in [Figure 1].
Figure 1: Active case search work model used for the active surveillance in all the wards across Lagos State in April 2020

Click here to view

The diagnosis of COVID-19 was confirmed by a respiratory sample testing positive for SARS-CoV-2 using a laboratory-based reverse transcription (RT)-PCR.

Identification of clusters

Communities were included based on the alerts from community members and the vulnerability of the communities as well as facilities with reported outbreaks through the rumour alert system set up by the state. The health workers moved from house to house to administer an electronic checklist. All members of the households of the selected community were eligible. Household members who declined to participate were excluded from the study.

Collection of case data and assessment of exposure

A standardised interviewer-administered electronic investigation form was used to collect sociodemographical information, clinical details and history for each possible case within the ward. Possible cases identified and reported were referred to one of the nine isolation centres designated for the management of patients with COVID-19 in Lagos. The activities from 14 days before symptom onset until isolation of suspected cases were mapped out, and their close contacts traced to identify the possible source of exposure. Contact tracing after symptom onset until isolation was done to identify case contacts for quarantine to break the transmission chain. Both these approaches were part of the containment strategy. All close contacts that have active symptoms were tested.

  Results Top

A total number of 12,739 participants were interviewed. The most common symptoms were fever (45.9%), general weakness (24.7%) and cough (20.4%). There was a male predominance of 57% [Table 1]. The overall mean age was 39.19 (standard deviation ± 17.75). The age group ≤30 years had the highest proportion of cases (34%), whereas the lowest proportion of cases was found amongst those of age group >60 years (11%), as shown in [Table 1]. Analysis of the data of the 11,828 persons for whom PCR results were available enabled the estimation of the validity of symptomatology.
Table 1: Demographic characters of the cases screened by symptoms suggestive of coronavirus disease 2019 during the active case search in the twenty local government areas, Lagos State in April 2020

Click here to view

Clinical presentations

The predominant symptoms that the screened persons presented with were fever (45.9%), general weakness (24.7%), cough (20.4%), sore throat (8.3%) and difficulty in breathing (6.3%), as shown in [Table 2].
Table 2: The frequencies of main symptoms as seen amongst persons screened during the active case search for coronavirus disease 2019 in the twenty Local Government Areas, Lagos State in April 2020

Click here to view

Based on the data of screened individuals during the active case search, fever, general body weakness and Cough were the most common symptoms reported amongst, and these were then validated against the results of RT-PCR, a gold-standard [Table 3]. An overall prevalence of 0.8% (n = 92) based on PCR positivity was recorded. For the different symptoms, negative predictive values ranged between 98.9% and 99.2%, whereas the positive predictive values ranged between 0.3% and 0.9% for different symptoms. In addition, the specificity and sensitivity values of these symptoms ranged from 50.2% to 93.2% and 6.5% to 28.3%, respectively.
Table 3: Validity of the presence of individual symptoms with polymerase chain reaction

Click here to view

As shown in [Table 3], where the validity of individual symptoms was considered, even fever that had the highest predictability had only a sensitivity of 28%, meaning it could correctly predict the infection in only 28 out of a hundred PCR-positive cases. This would result in a large number of false negatives, which could accelerate the epidemic if that diagnostic algorithm was adopted. PCR positivity levels were also very low even when combinations of symptoms were considered suggesting very poor yield when symptoms are used as a screening algorithm before PCR testing [Table 4].
Table 4: polymerase chain reaction positivity rates for combinations of symptoms

Click here to view

  Discussion Top

The magnitude of the impact of COVID-19 in developing countries is still unravelling amidst challenges for testing. Syndrome algorithms were used in the early days of the HIV/AIDS pandemic when testing was a challenge in many developing countries, and this study wanted to see to what extent symptoms could predict the presence of COVID-19. The complete clinical manifestation of the disease is not yet clear, and previously reported symptoms correspond with findings from this study such as fever, general weakness and cough.[13],[14],[15],[16] Differences in severity of presentation and outcome between Europe and Africa have been reported, but the reason for this is not known; speculations have been on the immunity, genes, climate and the young age of a majority of the population.[17],[18] Similar to what was observed in this study, the most common symptoms were fever, cough, general body weakness and complicated dyspnoea, whereas less commonly reported symptoms included headache, diarrhoea, haemoptysis, runny nose and phlegm-producing cough.[15],[19],[20]

Nigeria is in the early expanding transition period of the demographic transition theory, and a majority of the population are adolescents and young adults.[21] Therefore, this may explain why the probable cases are more within the younger age group, and in turn, why there is a greater prevalence of mild-to-asymptomatic cases across Lagos State. The absence of symptoms in a majority of infected persons makes the use of symptoms in predicting an infection a not too useful task. A similar study showed the prevalence of mild cases amongst younger age groups,[22],[23] and this may also explain the low mortality and reduced severity experienced so far in this outbreak.[8]

The control for the COVID-19 outbreak depends on case-based measures such as active and passive case ascertainment, effective contact tracing, rapid testing and diagnoses. The RT-PCR is regarded as a gold standard test for the molecular diagnosis of viral infections. In this study, the low positive predictive values for different symptoms reflect the limited effectiveness of predicting COVID-19 disease using symptoms. Even when symptoms were combined, there was no improvement in the ability of symptoms to predict COVID-19. It appeared that the predictive ability declined with increasing numbers of symptoms combined [Table 4].

This indicates that many of the symptom positive results from the screening during the active case search, before the RT-PCR test, were false positives. The negative predictive values, on the other hand, were high, reflecting reliability in predicting the absence of COVID-19 when no symptom exists. This finding is, however, to be taken with caution as a majority of COVID-19 cases can present asymptomatically.

A potential limitation of the study was the use of symptoms for the COVID-19 screening exercises, as asymptomatic cases would have been missed, thereby underestimating the prevalence of COVID-19 disease in the LGAs assessed. Furthermore, fever as a symptom is common in some other diseases such as malaria that are endemic in Nigeria and therefore may not be a useful predictive symptom in COVID-19 unless the community prevalence of the disease is very high. Future studies can assess the predictive value of other early symptoms such as sudden smell loss and nasal congestion, which have been reported to be significant indicators of COVID-19 disease in other populations[24].

  Conclusion Top

Symptoms should be recognised as a pointer to the disease and should raise a high index of suspicion. It is unlikely that symptoms alone will suffice to predict COVID-19 in patients. An additional measure, such as a confirmatory test by RT-PCR, is necessary to confirm positivity.

Financial support and sponsorship


Conflicts of interest

There are no conflicts of interest.

  References Top

2019 Novel Coronavirus (2019nCoV): Strategic Preparedness and Response Plan; 3 February, 2020. Available from: https://www.who.int/docs/default-source/coronaviruse/srp-04022020.pdf. [Last accessed on 2020 May 27].  Back to cited text no. 1
COVID-19 Dashboard by the Center for Systems Science and Engineering (CSSE) at Johns Hopkins University (JHU). Available from: https://coronavirus.jhu.edu/map.html. [Last accessed on 2020 Jul 08].  Back to cited text no. 2
Nigeria Center for Disease Control. Available from: http://covid19.ncdc.gov.ng. [Last updated on 2020 Jul 08; Last accessed on 2020 Jul 08].  Back to cited text no. 3
Cascella M, Rajnik M, Cuomo A, Dulebohn SC, Di Napoli R. Features, Evaluation, and Treatment of Coronavirus (COVID-19). 2020. In: StatPearls [Internet]. Treasure Island (FL): StatPearls Publishing 2020. PMID: 32150360. [https://pubmed.ncbi.nlm.nih.gov/32150360/].  Back to cited text no. 4
Huang C, Wang Y, Li X, Ren L, Zhao J, Hu Y, et al. Clinical features of patients infected with 2019 novel coronavirus in Wuhan, China. Lancet 2020;395:497-506.  Back to cited text no. 5
Imam Z, Odish F, Gill I, O'Connor D, Armstrong J, Vanood A, et al. Older age and comorbidity are independent mortality predictors in a large cohort of 1305 COVID-19 patients in Michigan, United States. J Intern Med 2020;469-76.  Back to cited text no. 6
Williamson EJ, Walker AJ, Bhaskaran K, Bacon S, Bates C, Morton CE, et al. OpenSAFELY: factors associated with COVID-19 death in 17 million patients. Nature 2020;584:430-6.  Back to cited text no. 7
Yang J, Zheng Y, Gou X, Pu K, Chen Z, Guo Q, et al. Prevalence of comorbidities and its effects in patients infected with SARS-CoV-2: A systematic review and meta-analysis. Int J Infect Dis 2020;94:91-5.  Back to cited text no. 8
Abstract of Local Government Statistics 2016. Lagos Bureau of Statistics Ministry of Economic Planning and Budget, Secretariat. Alausa, Ikeja: 2016.  Back to cited text no. 9
Nigeria Census-Nigeria Data Portal. Available from: https://nigeria. Opendataforafrica.org/xspplpb/nigeria-census. [Last accessed on 2020 May 27].  Back to cited text no. 10
Maduka O, Akpan G, Maleghemi S. Using Android and Open Data Kit Technology in Data Management for Research in Resource-Limited Settings in the Niger Delta Region of Nigeria: Cross-Sectional Household Survey. JMIR Mhealth Uhealth 2017;5:e171.  Back to cited text no. 11
Kit O. Open Data Kit Documentation; 2019. Available from http://www.academia.edu/download/59226450/ODK-Documentation20190512-81717-tt3 hwk.pdf. [Last accessed on 2020 May 25].  Back to cited text no. 12
Chen Y, Chen L, Deng Q, Zhang G, Wu K, Ni L, et al. The presence of SARS-CoV-2 RNA in the feces of COVID-19 patients. J Med Virol 2020;92:833-40.  Back to cited text no. 13
Chen H, Guo J, Wang C, Luo F, Yu X, Zhang W, et al. Clinical characteristics and intrauterine vertical transmission potential of COVID-19 infection in nine pregnant women: A retrospective review of medical records. Lancet 2020;395:809-15.  Back to cited text no. 14
Li Q, Guan X, Wu P, Wang X, Zhou L, Tong Y, et al. Early transmission dynamics in Wuhan, China, of novel coronavirus – Infected pneumonia. N Engl J Med 2020;382:1199-207.  Back to cited text no. 15
Africa Centre for Disease Control Dashboard. Available from: https://africacdc.org/covid-1/. [Last accessed on 2020 Jul 08].  Back to cited text no. 16
Dyer O. Covid-19: Africa records over 10000 cases as lockdowns take hold. BMJ 2020;369:m1439.  Back to cited text no. 17
Bowale A, Abayomi A, Idris J, Omilabu S, Abdus-Salam I, Adebayo B, et al. Clinical presentation, case management and outcomes for the first 32 COVID-19 patients in Nigeria. Pan Afr Med J 2020;35:24.  Back to cited text no. 18
Guan W, Ni Z, Hu Y, Liang W, Ou C, He J, et al, Clinical Characteristics of Coronavirus Disease 2019 in China, N Engl J Med 2020;382:1708-20.[DOI: 10.1056/NEJMoa2002032].  Back to cited text no. 19
Del Rio C, Malani PN. COVID-19 – New Insights on a Rapidly Changing Epidemic. JAMA 2020;323:1339-40.  Back to cited text no. 20
Akokuwebe ME, Okunola RA. Demographic transition and rural development in Nigeria. Dev Ctries Stud 2015;5:90-102.  Back to cited text no. 21
Dong Y, Mo X, Hu Y, Qi X, Jiang F, Jiang Z, et al. Epidemiological characteristics of 2143 pediatric patients with 2019 coronavirus disease in China. Pediatrics 2020;145:e20200702.  Back to cited text no. 22
Zhang C, Gu J, Chen Q, Deng N, Li J, Huang L. Clinical and epidemiological characteristics of pediatric SARS-CoV-2 infections in China: A multicenter case series. PLoS Med 2020;17:e1003130.  Back to cited text no. 23
Haehner A, Draf J, Dräger S, de With K, Hummel T. Predictive Value of Sudden Olfactory Loss in the Diagnosis of COVID-19. ORL J Otorhinolaryngol Relat Spec 2020;82:175-80.  Back to cited text no. 24


  [Figure 1]

  [Table 1], [Table 2], [Table 3], [Table 4]


Similar in PUBMED
   Search Pubmed for
   Search in Google Scholar for
 Related articles
Access Statistics
Email Alert *
Add to My List *
* Registration required (free)

  In this article
Article Figures
Article Tables

 Article Access Statistics
    PDF Downloaded192    
    Comments [Add]    

Recommend this journal