Accuracy of subjective assessment of fever by Nigerian mothers in under-5 children.

BACKGROUND: Many mothers still rely on palpation to determine if their children have fever at home before deciding to seek medical attention or administer self-medications. This study was carried out to determine the accuracy of subjective assessment of fever by Nigerian mothers in Under-5 Children. PATIENTS AND METHODS: Each eligible child had a tactile assessment of fever by the mother after which the axillary temperature was measured. Statistical analysis was done using SPSS version 19 (IBM Inc. Chicago Illinois, USA, 2010). RESULT: A total of 113 mother/child pairs participated in the study. Palpation overestimates fever by 24.6%. Irrespective of the surface of the hand used for palpation, the sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) of tactile assessment were 82.4%, 37.1%, 51.9% and 71.9%, respectively. The use of the palmer surface of the hand had a better sensitivity (95.2%) than the dorsum of the hand (69.2%). The use of multiple sites had better sensitivity (86.7%) than the use of single site (76.2%).
CONCLUSION: Tactile assessment of childhood fevers by mothers is still a relevant screening tool for the presence or absence fever. Palpation with the palmer surface of the hand using multiple sites improves the reliability of tactile assessment of fever.

INTRODUCTION

Fever is a common reason for seeking medical attention and accounts for more than 20% of emergency department visits.1 Many mothers and caregivers have perceived phobia associated with their child's fever for reasons that include serious infection, seizure, brain damage and death.2 Paediatricians have regularly advised parents to seek immediate medical attention if their children developed a fever because fever is of great diagnostic importance especially in developing countries where infectious diseases such as malaria and pneumonia are very prevalent.3

In developing countries, because of the relatively low level of literacy and financial constraint of procuring reliable thermometers, many parents rely on palpation to assess if their children have fever.4 The objective use of the thermometer to detect the presence of fever is often limited to health care settings in most resource-poor settings. An accurate determination of the absence of fever in a child assures parents and saves cost by preventing unnecessary investigations and medication.

Ownership of thermometer seems poor, although no study in our environment was found to support that. In a study in New York, USA, of the 78% of caregivers that owned thermometers, nearly half (48%) still utilised tactile method to assess fever.4 In another study, in India in contrast, only 15% owned thermometer with 23.8% of them not knowing how to use it.5 This underscores the fact that the use of palpation as a means of assessing fever by mothers is common and universal.

Majority of mothers and care givers in our environment still utilise tactile assessment to determine if their children have fever. Controversies over the reliability of tactile assessment of fever have been variously reported; while a study found it accurate enough, at least in younger infants,4 others reported that it overestimates the presence of fever.67 The reliability of the use of palpation to detect fever still remains uncertain.

This study was carried out to determine the reliability of tactile assessment of fever by Nigerian mothers in under-5 children and also to determine if the surface of the hand used influence the accuracy. The study further ascertained if palpation of a single or multiple anatomical sites influence the accuracy of tactile assessment of fever.

PATIENTS AND METHODS

Children under the age of 5 years presenting for care with their mothers were recruited for the study. Informed consent was obtained from their mothers before recruitment. Ethical approval was obtained from the research and ethics committee of Federal Medical Centre Owerri, Imo state Nigeria before commencement of the study.

The subjects were recruited from the children outpatient clinic and children emergency room of the hospital. Information such as age and sex of the child, mother's age and highest educational qualification (HEQ) were captured using a profoma. Thereafter, each eligible child had a tactile assessment of fever by the mother. The surface of the hand used and site/sites of palpation were noted. The response of the mother as regard presence or absence of fever was documented following which the child's axillary temperature was measured and documented. The axillary temperature was measured using mercury in glass thermometer, which was left in the axilla for 5 minutes before reading.

Statistical analysis was done using SPSS version 19 (IBM Inc. Chicago Illinois, USA, 2010). Frequencies, percentages and mean were analysed where appropriate. The sensitivity, specificity, positive predictive values (PPVs) and negative predictive values (NPVs) of tactile assessment was calculated using axillary temperature of ≥37.5°C as cut-off for fever. The different hand surfaces used for palpation, as well single versus multiple sites of palpation were also compared.

RESULTS

A total of 113 mother/child pairs participated in the study. The children were aged 2 days to 59 months. There were 60 males and 53 females giving a male: Female ratio of 1.1:1. All the mothers had some form of formal education. Seven (6.2%) of the mothers had only primary school education, while 38 (33.6%) and 68 (60.2%) had secondary and tertiary education, respectively. Fever detected by touch was reported in 81 (71.6%) children by mothers while, only 65 (57.5%) mothers were able to correctly predict the presence of fever in their children. This implies that mothers are likely to overestimate the presence of fever by 24.6%. There was no statistical significant difference in the ability of the mothers to correctly predict fever with respect to their HEQ (χ2 = 2.66, P = 0.265). This is shown in Table 1.

Table 1

Comparison of fever detection by palpation with the mothers’ highest educational qualification (HEQ)

Irrespective of the part of hand used for palpation, the sensitivity of detection of fever by palpation was 82.4%, while the specificity was 37.1%. The use of the palmer surface of the hand had a better sensitivity (95.2%) than the dorsum of the hand (69.2%). Table 2 shows the sensitivity, specificity, PPV and NPV of the part of hand used for palpation.

Table 2

Sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) of the part of hand used for palpation

A total of 71 (62.8%) of the mothers palpated at more than one site while the remaining 42 (37.2%) used only single site. Table 3 shows the different sites used by the mothers in palpation, while Table 4 compares the sensitivity, specificity, PPV and NPV of use of single site versus multiple sites in palpation. The use of multiple sites had better sensitivity (86.7%) than the use of single site (76.2%).

Table 3

Sites used by the mothers to palpate for fever

Table 4

Palpation of single site versus multiple sites

DISCUSSION

The present study shows that mothers can be trusted to a good extent to assess for the presence of fever in children by palpation. A sensitivity of 82.4% obtained in this study is similar to those reported by Okposio and Abhulimhen-Iyoha in Benin, Nigeria as well as Graneto and Soglin in Chicago, USA, who documented sensitivity of 89.2% and 84%, respectively.89 Some studies in Nigeria have reported a higher sensitivity value of 96.3% and 94.6%, respectively.610 A specificity of 37.1% obtained in the present study is quite poor and is lower than the specificities reported by other workers ranging from 64.3% to 82%.489 It is, however, higher than 23% reported by Akinbami et al.6 The heterogeneity in the methodology of these studies may have been responsible for the differences obtained between the present study and some of the earlier studies. For instance, Akinbami et al.,6 used rectal temperature of ≥38°C as the cut-off of fever, Wammanda and Onazi10 adopted axillary temperature of ≥37.2°C as the cut-off for fever, while the present study and that of Okposio and Abhulimhen-Iyoha8 used axillary temperature of ≥37.5°C as the definition of fever. Although the specificity in the present study was poor, the relatively high sensitivities obtained in all the studies still make maternal palpation for the presence of fever a considerably useful method of assessment of childhood fever. Although, palpation overestimates fever, the elicitation of fever by mothers would encourage early presentation as it has been documented that the concern for fever is a strong motivational factor for seeking medical care.2

The educational qualification of a mother did not influence the accuracy of palpation for the presence of fever. It is therefore expected that any mother, irrespective of her educational background should be able to appreciate fever by palpation. With a NPV of 71.9%, if a mother says her child does not have fever, her history of subjective assessment of fever without the use of a thermometer is about 72% reliable. The likelihood of a child adjudged to be febrile by a mother to be truly febrile measured by the PPV in the present study is 51.9%. This is higher than 39% reported by Wheybrew and co-workers.7 This higher PPV reported in this study may have been influenced by the prevalence of fever in the present study. In the present study, 57.5% of the study subjects had fever, while in the study of Wheybrew et al.,7 27% of children had fever.

The palmer surface of the hand was the most commonly used hand surface by mothers to palpate for fever. The sensitivity, specificity, NPV and PPV of palmer surface was higher than that of the dorsal surface or the use of both surfaces. This may be attributed to the fact that receptors for detection of heat and temperature are more concentrated on the palmer surface and finger tips.11 The use of both surfaces may confuse the mother as the sensation obtained from both are likely to vary, hence the reduced sensitivity obtained in this study when compared with palmer surface alone.

The forehead was the most common site used for tactile assessment of fever in this study. This corroborates reports from previous studies, which also found the forehead to be the most common anatomical site used for tactile assessment of fever.4 This may possibly be explained by the fact that the head is easily accessible compared with the other parts of the body.

It was observed that palpation of multiple sites was more sensitive than use of single site. This is in agreement with an earlier report by Okposio and Abhulimhen-Iyoha as well as Singhi and Sood.812 This implies that if fever is to be assessed by palpation without the use of a thermometer, palpation of multiple body sites should be utilised.

CONCLUSION

Tactile assessment of fever in children by mothers is still a sensitive screening method. Palpation with the palmer surface of the hand using multiple sites improves the reliability of tactile assessment of fever. Clinicians should not discountenance mothers’ complaint of fever in their children detected by palpation without the use of a thermometer. Mothers should be taught to palpate at multiple sites with the palmer surface when palpating for the presence of fever in their children to improve the accuracy.