Efficacy of Canadian computed tomography head rule in predicting the need for a computed-axial tomography scans among patients with suspected head injuries.

CONTEXT: The use of imaging modalities is crucial in the diagnostic field of critical medicine. However, the ethical and economic use of these techniques has become a major concern especially in resource-poor settings. The Canadian computed tomography Head Rule (CCHR) is being increasingly used all over the world to evaluate the necessity of a Computer-assisted Tomography (CT) scan in patients with suspected head injury. AIM: The aim of the current study is to evaluate the efficacy of CCHR to predict the occurrence of head injury, as evidenced radiologically by a CT Head, at a government tertiary care clinical setting in south India. SETTING AND
DESIGN: The design was that of a hospital-based cross-sectional survey conducted at the Medical College Hospital, Thiruvananthapuram (Kerala, India).
MATERIALS AND METHODS: The study subjects were patients with suspected head injury evaluated at the Surgical Casualty Department of the study setting. Fifty consecutive patients with suspected head injury were enrolled in the study. STATISTICAL ANALYSIS: The Chi-square test was used to assess the statistical significance of association between the outcome variable and the exposure characteristics. The diagnostic ability of the Glasgow Coma Scale (GCS) and CCHR were expressed in terms of sensitivity and specificity by considering CT diagnosed Head injury as the gold standard diagnostic tool.
RESULTS: Clinical manifestations as measured by a GCS score < 13 failed to significantly predict a head injury in the CT scan. However, the same became statistically significant when the CCHR was added to the GCS score as a predictor (P value < 0.001). The sensitivity of the tool in predicting a head injury rose from 23.3 to 96.7%.
CONCLUSION: The current study suggested that the CCHR could act as an excellent decision rule to indicate the need of a CT scan. The need of a decision rule was warranted in the context of the growth of newer diagnostic imaging facilities in India.

INTRODUCTION

Injuries are emerging as a major public health problem and Road Traffic Accidents are one of the leading causes of mortality the world over. India, with its rapid socioeconomic development is undergoing industrialization and motorization at a fast pace. As its price, injuries in the country are clearly on the rise and the economic burden that they cause is formidable. The newer diagnostic facilities, especially the imaging techniques, have undergone a revolutionary uplift since the 1990s, in India.[1] Radiological investigations form an inevitable component in the work-up of a victim of head injury. The out-of-pocket health expenditure for radiological investigations, like a CT scan, is tremendous. The irrational and unethical test-ordering behavior of the care providers worsens the situation. Commensurately, as these services are not widely available in the public sector, it urges the private diagnostic facilities to provide low quality services and to incur a high cost for these investigations.[1] Lack of definite standards and protocols for advising a CT scan leads to clinical uncertainty among the doctors, which results in inadvertent ordering of these investigations.

A head injury can present with a multitude of symptoms, but it is difficult to predict head injury with these symptoms. The positive predictive value of a single symptom like vomiting alone is less than 5%.[2] Several other signs and symptoms like posttraumatic amnesia, loss of consciousness, posttraumatic seizure, headache, and focal neurological deficit, are indicative of an underlying head injury. An accompanying skull fracture, elderly patient of age more than 60 years, bleeding disorders or treatment with anticoagulants has got an increased risk of head injury.[3] Clinicians widely rely on the GCS score in predicting head injury. However, a low GCS score does not always accurately predict the outcome of severe traumatic brain injury.[4] The Canadian CT Head Rules (CCHR) consist of a set of clinical signs and conditions and is being increasingly used all over the world in various settings[56] as a clinical guide, to decide on the usage of a CT scan. This conclusion has been drawn from the results of prospective cohort studies conducted in the Emergency Departments of ten large Canadian Hospitals, with an objective to develop a protocol to be followed before advising a CT scan, for a patient with a minor head injury.[7] They derived a CT head rule, according to which a patient suspected of having a minor head injury during the past 24 hours must have any one of the five high risk factors like failure to reach a GCS of 15 within two hours, suspected open skull fracture, any sign of basal skull fracture, vomiting of more than two episodes, and age above 65 years. The two additional medium-risk factors were amnesia before impact for more than 30 minutes, and a dangerous mechanism of injury.[7] The CCHR is being increasingly used in some of the North American and European regions. In a low resource country like India, it is important to have a rational approach to the use of these diagnostic facilities. The aim of the current study is to evaluate the efficacy of CCHR to predict the occurrence of a head injury, as evidenced radiologically by a CT Head, at a government tertiary care clinical setting in south India.

MATERIALS AND METHODS

This hospital-based, cross-sectional survey was conducted at the Government Medical College Hospital (MCH), Thiruvananthapuram, Kerala, India. This is the first medical college to be set up in the state, at its capital city, Thiruvananthapuram, and is one of the biggest tertiary care hospitals in Kerala, with a very high patient turn-over rate. The people belonging to the poor income group often depend on the public health delivery system because of the reduced healthcare expenditure in this hospital for both diagnostic and therapeutic services.

The study subjects were patients with suspected head injury evaluated at the surgical casualty department of the study setting. Fifty consecutive patients were enrolled in the study during the study period of one month. A structured questionnaire was used for data collection, which included questions related to the sociodemographic, clinical, and radiological features. All items in the GCS and CCHR were included in the data collection tool. A written informed consent in the local language was obtained from the study subjects or their immediate relatives (if the victim was not in a position to take a decision because of the injury) before the administration of the study tool.

Statistical analysis

A lesion indicating head injury in the CT scan was the major outcome variable. All the categories of the covariate list were represented in percentages. The Chi-square test was used to assess the statistical significance of association between the outcome variable and the exposure characteristics. A significance level of 95% and a power of 80% were assumed in all statistical analyses. The diagnostic ability of GCS and CCHR were expressed in terms of the sensitivity and specificity, by considering CT-diagnosed head injury as the gold standard diagnostic tool. Statistical Package for Social Sciences (SPSS) version 16 was used for data analysis.

RESULTS

Thirty out of 50 patients (60%) had radiological evidence of head injury. The age of the victims of injury ranged between 13 and 72 years and a majority (29, 58%) of them were in the age range of 25 to 40 years. Men made up 88% of the study subjects. Manual laborers (15, 30%), drivers (10, 20%), and painters (5, 10%) were found to be more prone to head injury. A majority (31, 62%) of the patients admitted with head injury had an educational qualification of below tenth standard. This could be due to the fact that these patients hailed from the low income group, who sought health care from a government medical college hospital.

The prediction of the GCS score less than 13 left out a majority of the cases with head injury as evidenced by the results of the CT scan [Table 1]. However, combining the CCHR with the GCS score to predict head injury significantly raised the number of cases correctly diagnosed and was almost comparable to the results generated by the CT scan (P value < 0.001). Out of the nine individuals who had a GCS score of less than 13, seven (positive predictive value = 77.8%) had a head injury in the CT scan. However, out of the 30 individuals who actually suffered from head injury according to the gold standard, only seven (Sensitivity = 23.3%) were detected by very low GCS [Table 1]. The sensitivity of the tool rose from 23.3 to 96.7% by predicting 29 out of the total 30 individuals who were true positives [Table 1]. Only one case of head injury was missed by the CCHR and it was of a 45-year-old man diagnosed with having a mild cerebral contusion in the CT Head. Out of the 35 individuals predicted to be having head injury by the new criteria, 29 (positive predictive value = 82.9%) had radiological evidence of head injury. The negative predictive value also rose from 43.9 to 93.3%. However, there was a mild reduction of 10% noted in the specificity [Table 1].

Table 1

Comparison of glasgow coma score (< 13) and canadian CT head rule in a lesion in CT scan of the Head

DISCUSSION

The predictive power of CCHR to pinpoint a high-risk candidate for a CT Scan Head appeared to be high in the current study. The positive likelihood ratio (+ LR = Sensitivity / 1-specificity) of a diagnostic test was regarded as the best measure of diagnostic accuracy of a test.[8] The + LR for CCHR appeared to be 3.22, which was distinctly higher than that of the GCS alone, which was 2.32. The negative likelihood ratios were 0.05 and 0.85, respectively. A higher value of + LR and lower value of - LR suggested the usefulness of the criterion as a diagnostic algorithm.[89] The higher values of the positive predictive value and negative predictive value for CCHR showed that this criterion was very useful in the setting where the test was performed. The current study, even though limited by its small sample size, suggested that the CCHR could act as an excellent decision rule to indicate the need for a CT scan.

Indian healthcare facilities have succeeded in gaining significant attraction in the international market and are undergoing transformation rapidly. In India, documented evidence suggests that the demand and supply of imaging services have been increasing since the last two decades.[1] In this context, the irrational ordering of imaging modalities has become the trend in most parts of the country. Implementation of a guideline, which can correctly predict head injury in healthcare institutions, can go a long way in easing the economic burden of the patients.

A large cohort study, conducted in USA, suggested that the number of cranial CT scans conducted among pediatric patients could have been reduced by adopting the rule, which would ultimately increase patient safety, while reducing medical expense.[5] Several other studies have also documented the usefulness of CCHR in predicting head injuries in western emergency care settings.[5-710-13] Compared to other decision rules to predict a head injury, CCHR appears to be more useful and accurate in ruling out the need for a CT scan.[10] Results of the current study also point in the same direction, with a sharp and significant increase in sensitivity to 96% using CCHR as compared to 23% using GCS. In terms of diagnostic efficacy it appears to have a comparable sensitivity and more specificity than the major competing diagnostic algorithm, known as the New Orleans Criteria.[11] The usefulness of CCHR has been studied in many settings. A large cohort study conducted in USA suggests that a significant number of cranial CT scans conducted among pediatric patients may have been reduced by adopting the rule in their setting.[12] However, it is not widely being practiced even in the Emergency Departments of Canada, from where the tool took birth.[13] Although an economic analysis has not been intended in this study, it is evident that 30% of the victims would have been exempted from a CT scan Head if CCHR were followed. In this setting, to conduct a CT scan on 100 patients it would amount to nearly 1200 US dollars. If the radiation exposure also gets accounted for, then the overall cost would flare up even further.

CONCLUSION

A cost-effective clinical diagnostic strategy is being called for by the significant increase in the burden of head injuries in developing countries during the past few decades. This study brings to light the effectiveness of a clinical tool, CCHR, for predicting head injury among those suspected with it, which is rather underutilized worldwide. The ambit of this tool is vast, especially in a resource-poor setting like India, where there is a huge burden on the patient's own pocket, of health care expenditure. Selectively subjecting those who are at high and medium risk, according to the CCHR, is considered to be optimal, rather than subjecting all patients to the scan or not investigating them at all.[14] The advantage in terms of the cost will be huge when extrapolated to a highly populous country like India, where millions of people are undergoing diagnostic imaging techniques every year. Furthermore, the high sensitivity and positive predictive value of the tool suggest the incisive epidemiological impact of the tool. The adoption of such guidelines will reduce medical expenses and the radiation dose, thereby increasing patient safety.