Journal of Postgraduate Medicine
 Open access journal indexed with Index Medicus & EMBASE  
     Home | Subscribe | Feedback  

[Download PDF
Year : 2012  |  Volume : 58  |  Issue : 4  |  Page : 265-269  

Evaluation of clinical features scoring system as screening tool for influenza A (H1N1) in epidemic situations

P Ranjan1, A Kumari2, R Das3, L Gupta1, SK Singh4, M Yadav1,  
1 Department of Medicine, Lady Hardinge Medical College and Associated Hospitals, New Delhi, India
2 Department of Obstetrics and Gynaecology, Lady Hardinge Medical College and Associated Hospitals, New Delhi, India
3 Department of Community Medicine, Lady Hardinge Medical College and Associated Hospitals, New Delhi, India
4 Airport Health Organization, Indira Gandhi International Airport, New Delhi, India

Correspondence Address:
S K Singh
Airport Health Organization, Indira Gandhi International Airport, New Delhi


Background: Influenza A (H1N1) hit the headlines in recent times and created mass hysteria and general panic. The high cost and non-availability of diagnostic laboratory tests for swine flu, especially in the developing countries underlines the need of having a cheaper, easily available, yet reasonably accurate screening test. Aims: This study was carried out to develop a clinical feature-based scoring system (CFSS) for influenza A (H1N1) and to evaluate its suitability as a screening tool when large numbers of influenza-like illness cases are suspect. Settings and Design: Clinical-record based study, carried out retrospectively in post-pandemic period on subject«SQ»s case-sheets who had been quarantined at IG International Airport«SQ»s quarantine center at Delhi. Materials and Methods: Clinical scoring of each suspected case was done by studying their case record sheet and compared with the results of RT-PCR. RT-PCR was used to confirm the diagnosis (Gold Standard). Statistical Analysis: We calculated sensitivity, specificity, positive and negative predictive values of the clinical feature-based scoring system (the proposed new screening tool) at different cut-off values. The most discriminant cut-off value was determined by plotting the ROC curve. Results: Of the 638 suspected cases, 127 (20%) were confirmed to have H1N1 by RT-PCR examination. On the basis of ROC, the most discriminant clinical feature score for diagnosing Influenza A was found to be 7, which yielded sensitivity, specificity, positive, and negative predictive values of 86%, 88%, 64%, and 96%, respectively. Conclusion: The clinical features scoring system (CFSS) can be used as a valid and cost-effective tool for screening swine flu (influenza A (H1N1)) cases from large number of influenza-like illness suspects.

How to cite this article:
Ranjan P, Kumari A, Das R, Gupta L, Singh S K, Yadav M. Evaluation of clinical features scoring system as screening tool for influenza A (H1N1) in epidemic situations.J Postgrad Med 2012;58:265-269

How to cite this URL:
Ranjan P, Kumari A, Das R, Gupta L, Singh S K, Yadav M. Evaluation of clinical features scoring system as screening tool for influenza A (H1N1) in epidemic situations. J Postgrad Med [serial online] 2012 [cited 2022 Jun 28 ];58:265-269
Available from:

Full Text


Swine flu was a major global pandemic of the century that hit the headlines in recent times. [1],[2] The fact that the swine flu (influenza A (H1N1)) virus, spread rapidly and caused mortality in many otherwise young and healthy adults, led to widespread panic among the masses. [3],[4],[5] Daily media reports of climbing mortality figures in the metropolises led to a situation of mass hysteria and widespread terror against this obnoxious pathogen. [6],[7],[8] As a result, people with common cold or influenza-like illness (ILI) started rushing to hospitals for costly RT-PCR tests and stored Oseltamivir and N 95 masks that were available in market at manifold enhanced prices. [9],[10],[11],[12]

The first case of the disease was reported from Mexico on 18 th March 2009, and it rapidly spread to the neighboring United States and Canada. [13],[14] Subsequently, the disease spread to all the continents. World Health Organization (WHO) had raised the level of influenza pandemic alert from phase 5 to 6 on 11 th June 2009. [6] The last update released by WHO on August 6, 2010 reported the spread of laboratory-confirmed cases of pandemic influenza H1N1 2009 in more than 214 countries, including more than 18,449 deaths. [15] India reported its first case on 16 th May 2009. [16] Most of the initial cases reported subsequently were those travelling to India from affected countries. Till 2010 December, 202,790 persons were tested for H1N1. Of these, 46,131 (23%) reported positive and rest were negative. [17] The mortality of the disease was 2,728 (6%). [18] In due course of time, the infection and disease was brought under control, and W.H.O. announced H1N1 influenza to be in "post-pandemic phase" on August 10, 2010. [19] It is expected that the influenza A (H1N1 2009) virus will circulate as seasonal influenza virus for some years to come and localized outbreaks of various magnitude can occur. [20]

It is truly said that pandemics are lived forward but are understood backwards. [21] It is now believed that the progression/spread of swine flu pandemic (H1N1 2009) could have been halted far before if quarantine of cases had been practiced effectively in early stages. [22],[23] This would have been possible by early screening of cases who otherwise kept on spreading the infection. Unfortunately, the diagnosis of swine influenza (by viral RNA estimation with RT-PCR) is a costly affair and is not available everywhere, especially in the developing world. [24] We, therefore, attempted to develop a clinical feature scoring system (CFSS) and evaluated its validity that could be available readily as cost-effective alternative. This screening tool (i.e. CFSS) holds its significance even in the post-pandemic phase of swine flu (H1N1) epidemic as a part of our preparedness for subsequent outbreaks that may occur by further mutation in this highly unstable virus. Currently, the H5N1 avian influenza strain is in pandemic phase 3, hence causing sporadic human infections. In the next few years, it may progress to higher phases in the pandemic classification. [25],[26]

 Materials and Methods

The present study is a case record-based study done retrospectively on the case record sheets of 638 influenza A (H1N1) suspects who had been quarantined at the International Airport between 25 th April to 4 th December 2009. The study falls in the category of natural experiment, and all cases of the sampled universe have been studied. The power of the study, taking the 2-tailed Z value related to α=0.05 as 1.96 was derived to be 0.83. An ethical approval was given by the respective institutional ethics committees of the concerned institutions.

The suspect case identification, its documentation, and their final management was in accordance with the current guidelines suggested by WHO and MOHFW. In an effort to prevent the transmission of H1N1 infection in India, all international air passengers travelling to India during that period were screened at the screening health desk (prior to an immigration check) as per the guidelines of the Ministry of Health and Family Welfare, Government of India.

A "suspect" had been defined as an individual having acute febrile (fever >38°C), respiratory illness with history of recent close contact with a confirmed case of A (H1N1) (within 7 days), or had travelled to a community/was resident of a community with one or more confirmed cases. A total of 638 subjects had been identified as suspects, and all were referred to the Airport Quarantine Center compulsorily.

At the Airport Quarantine Center, each suspect had been evaluated thoroughly by a specialist physician and by a specially-trained microbiologist who collected 1 nasal and 1 oro-pharyngeal swab sample. Nasal swab was taken from anterior turbinate area, and oropharyngeal swab was taken from posterior wall of oropharynx. Sample swab was taken with a stick made-up of Dacron tip with plastic shaft, which was put in vials containing

2 ml of virus transport media. Samples were appropriately packaged and were then transported to the apex microbiological testing center in wet ice box. RT-PCR was done as per the methodology and guidelines recommended by WHO. The sample was considered positive for influenza A (H1N1) if it was positive for influenza A but negative for seasonal H1 and H3 viruses.

All information had been recorded on a standardized case record sheet, which served as the basis for this study [Table 1]. Our study methodology consisted of analyzing the case record sheets of all the 638 subjects who had been quarantined at the Airport Quarantine Center in 2009. A pooled analysis of all study subjects was initially done to characterize them. We next categorized our subjects into 2 groups, influenza A (H1N1) and non-A (H1N1) groups based on the RT-PCR report. Clinical scoring of all subjects was done on the basis of clinical feature-based scoring system that was developed previously by giving due weightage to fever, URTI and history of contact as explained in [Table 2].{Table 1}{Table 2}

During the epidemic, WHO and MOHFW had recommended categorization of influenza A (H1N1) cases into A, B, or C category (for identifying those requiring no intervention, home-based intervention or hospital-based intervention). Our CFSS is a modification and elaboration of this classification system. The CFSS took into its purview all signs/symptoms that had a reasonably high relative prevalence among cases vis-ΰ-vis controls (non-influenza A H1N1). We included all parameters that had been identified by expert clinicians who managed the cases during the pandemic and other health policy makers (from WHO and MOHFW) and accorded a weightage corresponding to its probability of being associated with influenza A (H1N1). We ensured blinding at all levels to minimize the possibility of 'observer bias'. The CFSS was developed by the investigators prior to the analysis of case record sheets. Subjects as well as specialist doctors and microbiologist were unaware of study methodology or results. Separate investigators were assigned to calculate scoring (on the basis of CFSS) and to record the result of RT-PCR.

We checked the ability of the scoring system to identify influenza A (H1N1) cases at varying values. This, we did by plotting a matrix of sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) for different CFSS values. Finally, we found a score of CFSS that could be used as a screening tool in epidemic situations, especially when RT-PCR was not available. This most discriminant CFSS cut-off value was identified by constructing a receiver-operator characteristic (ROC) curve. The value on the top left corner was taken for most discriminant value [Figure 1].{Figure 1}


A total of 638 passengers were identified as suspected cases of swine flu at the entry screening health desk located at arrival terminal of the Airport and referred to the quarantine center. Out of 638 suspected cases, 127 (19.9%) cases reported positive for H1N1 (swine flu) with RT-PCR test. General characteristics of the study population are listed in [Table 3]. The presentation of symptoms and signs of confirmed cases are listed in [Table 4].{Table 3}{Table 4}

Of all the confirmed cases of swine flu, high grade fever was present in 39%, moderate grade was present in 38% where as low grade fever was present in 23% of cases. Cough, sore throat, and running nose were the 3 most common symptoms present in 73%, 61%, and 59% cases, respectively. History of a confirmed case in community was found in 84% in the community, and only 39% of the cases had history of close contact.

Determination of the cut-off value for CFSS

We calculated sensitivity, specificity, positive, and negative predictive values of the clinical feature scoring system (new screening test) at different cut-off values, whereas RT-PCR was used to make diagnosis (Gold Standard). The results are tabulated in [Table 5].{Table 5}

On the basis of ROC [Figure 1], the most discriminant clinical feature score when taken as 7, yielded the best combination of sensitivity, specificity, positive and negative predictive values for the proposed screening tool. The cut-off value of 7 yielded a positive predictive value of 64% and a negative predictive value of 96% while maintaining sensitivity and specificity at 86% and 88%, respectively.


We carried out a double-blind observational study in which the case records of 638 suspected cases of swine flu (H1N1) were studied. Neither cases nor observer (specialists who prepared the case records) were aware of the study. After RT-PCR, 127 (19.9%) of the suspected cases were confirmed for swine flu. The case positivity rate in our study (19.9%) was approximately same as reported in Ministry of Health and Family Welfare Report (23%). [19] Fever was present in all suspected cases as inclusion criteria. Low grade fever was present in almost 1/4 th of cases. This supports the notion that many cases of swine flu present with mild symptoms. [27] In our study, fever, cough, and sore throat were the 3 most common symptoms present in the swine flu patients. This is in accordance with the previous studies done worldwide. Our clinical features-based scoring system was developed on the basis of the Guidelines of Ministry of Health and Family Welfare, Government of India. During the pandemic period, Ministry of Health and Family Welfare had issued guidelines, which categorized suspected cases of swine flu (H1N1) in category A, B, C based upon clinical manifestations. [28] Based on this, cases were screened for home isolation, testing, treatment, and hospitalization. Clinical feature scoring system was also used in some other parts of the world. In Mexico, a scoring system was used as Hospital Triage System for adult patients for complimenting clinician's judgment in treatment and hospitalization of cases of swine flu. [29]

Finding a cut-off value was very important part of the study. In [Table 5], we had a set of values of sensitivity and specificity at different cut-off values. Choosing lower cut-off value would have resulted in high sensitivity and low specificity (i.e. high false positivity rate) - leading to unnecessary anxiety, panic reaction, and overuse of oseltamivir, whereas, higher cut-off would have resulted in high specificity and low sensitivity (i.e. high false negativity rate). [30] Missing a case is equivalent to permitting further transmission of the virus in the community. Fortunately, the virus swine flu (H1N1) has been observed to have overall low mortality. [27] Thus, a good choice for a test cut-off value was the value, which corresponds to a point on the ROC curve nearest to the upper left corner of the ROC graph. Based upon these analyzes, the most discriminant cut-off value for CFSS was 7. At cut-off value of 7, sensitivity and specificity was 86% and 88% respectively with positive predictive value of 64% and a negative predictive value of 96%. As a screening tool, this clinical feature scoring system has got the advantage of being cost-effective, simple, easy to administer, and rapid. Besides, it fulfills the basic criteria of acceptability, validity, and repeatability with some scope of observer variation.

This study, to our knowledge, is the first of its kind that compared a clinical feature scoring system (CFSS) with RT-PCR for diagnosis for influenza A (H1N1). With sensitivity and specificity of nearly 90%, it can be a cheaper alternative for screening swine flu. It can also be used as for developing a triage system for treatment and hospitalization of cases during future pandemic of influenza A (H1N1).

Our study has certain limitations. Firstly, it enrolled only international airlines passengers whose profile is different from that of general population. Secondly, the CFSS was dependent on history given by subjects. Thirdly, clinical examination was done by different specialists deployed at different times, thus leaving the scope of inter-observer variations.

The lessons from history highlights the fact that studies of this kind are needed to devise the strategy for appropriate handling of future epidemics. Influenza virus is unique in its ability to undergo genetic reassortment with fresh outbreaks every few years. Moreover, even in post-pandemic period, local outbreaks of influenza A (H1N1) infection continue to occur, and disproportionately affect a younger age group with higher mortality. Our study, thus, holds significance and relevance and makes way for the use of proposed CFSS as cost-effective and readily-available screening tool for influenza A (H1N1) in epidemic situations.


1Al Hajjar S, McIntosh K. The first influenza pandemic of the 21st century. Ann Saudi Med 2010;30:1-10.
2Amato Gauci AJ, Zucs P, Snacken R, Ciancio B, Lopez V, Broberg E, et al. The 2009 A(H1N1) pandemic in Europe A review of the experience [Internet]. [cited 2011 Sep 25]; Available from:
3Ma J, Dushoff J, Earn DJ. Age-specific mortality risk from pandemic influenza. J Theor Biol 2011;288:29-34.
4Crum-Cianflone NF, Blair PJ, Faix D, Arnold J, Echols S, Sherman SS, et al. Clinical and epidemiologic characteristics of an outbreak of novel H1N1 (swine origin) influenza A virus among United States military beneficiaries. Clin Infect Dis 2009;49:1801-10.
5Adhisivam B, Kar SS, Sujatha S, Dutta TK. Trend of Morbidity Related to Pandemic Influenza A (H1N1) Infection in a Tertiary Care Teaching Institution in South India. The Indian Journal of Pediatrics [Internet]. 2011 Jun 25 [cited 2011 Sep 25]; Available from:
6Burns SM. H1N1 influenza is here. J Hosp Infect 2009;73:200-2.
7Brown ST, Tai JH, Bailey RR, Cooley PC, Wheaton WD, Potter MA, et al. Would school closure for the 2009 H1N1 influenza epidemic have been worth the cost?: A computational simulation of Pennsylvania. BMC Public Health 2011;11:353.
8Jehn M, Kim Y, Bradley B, Lant T. Community knowledge, risk perception, and preparedness for the 2009 influenza A/H1N1 pandemic. J Public Health Manag Pract 2011;17:431-8.
9Thomas C. N95 mask: No Stock in Bangalore [Internet]. MIW --- Men in Web. 2009 Aug 9;Available from:
10Times Now. H1N1 masks sold in black market [Internet]. [cited 2011 Sep 25]. Available from:
11Lochi L, Lamba L. Complete Information about N95 Mask for Protecting Swine Flu [Internet]. [cited 2011 Sep 25]; Available from:
12Express News Service. Mask effect: manufacturer stops supply to local chemists to avoid black marketing - Indian Express [Internet]. 2009 Aug 14 [cited 2011 Sep 25]; Available from:
13Perez-Padilla R, de la Rosa-Zamboni D, Ponce de Leon S, Hernandez M, Quiñones-Falconi F, Bautista E, et al. Pneumonia and respiratory failure from swine-origin influenza A (H1N1) in Mexico. N Engl J Med 2009;361:680-9.
14Schnitzler SU, Schnitzler P. An update on swine-origin influenza virus A/H1N1: A review. Virus Genes 2009;39:279-92.
15WHO,Pandemic (H1N1) 2009 [Internet]. WHO; 2010 [cited 2011 Sep 25]. Available from:
16Sharma S. First swine flu case in India [Internet]. 2009 May 16 [cited 2011 Sep 25]; Available from:
17Epidemiological Trends In India [Internet]. [cited 2011 Sep 25]; Available from:
18Clinical Profile and Outcomes of H1N1 patients admitted in a tertiary care hospital in Karnataka.pdf [Internet]. [cited 2011 Sep 25]; Available from:
19Margaret Chan. H1N1 in post-pandemic period [Internet]. WHO; 2010 [cited 2011 Oct 8]. Available from:
20Tian D, Wang Y, Zheng T. A novel strategy for exploring the reassortment origins of newly emerging influenza virus. Bioinformation 2011;7:64-8.
21Mahajan R, Grover A. H1N1 2009 influenza pandemic: Looking for a blessing in disguise. Int J App Basic Med Res 2011;1:3-4.
22Doshi P. The elusive definition of pandemic influenza. Bull World Health Organ 2011;89:532-8.
23Myers KP, Olsen CW, Gray GC. Cases of swine influenza in humans: A review of the literature. Clin Infect Dis 2007;44:1084-8.
24Leonardi GP, Mitrache I, Pigal A, Freedman L. Public Hospital-Based Laboratory Experience during an Outbreak of Pandemic Influenza A (H1N1) Virus Infections. J Clin Microbiol 2010;48:1189-94.
25Zündorf I, Dingermann T. [The new (threatening?) influenza pandemics]. Pharm Unserer Zeit 2011;40:120-5.
26Cumulative Number of Confirmed Human Cases of Avian Influenza A/(H5N1) [Internet]. Global Alert and Response. 2011 Aug 9 [cited 2011 Sep 25]; Available from:
27Fitzgerald DA. Human swine influenza A [H1N1]: Practical advice for clinicians early in the pandemic. Paediatr Respir Rev 2009;10:154-8.
28Clinical Management Protocol-Pandemic influenza A H1N1.pdf [Internet]. [cited 2011 Sep 25]; Available from:
29Rodriguez-Noriega E, Gonzalez-Diaz E, Morfin-Otero R, Gomez-Abundis GF, Briseño-Ramirez J, Perez-Gomez HR, et al. Hospital triage system for adult patients using an influenza-like illness scoring system during the 2009 pandemic--Mexico. PLoS ONE 2010;5:e10658.
30Simon D, Boring JR. Sensitivity, specificity, and predictive value [Internet]. In: Walker HK, Hall WD, Hurst JW, editors. Clinical methods: The history, physical, and laboratory examinations. 3rd ed. Boston: Butterworths; 1990 [cited 2011 Oct 8]. Available from: http://www.ncbi.nlm.nih .gov/books/NBK383/

Tuesday, June 28, 2022
 Site Map | Home | Contact Us | Feedback | Copyright  and disclaimer