Development and Validation of a Novel Method for Converting the Japan Coma Scale to Glasgow Coma Scale

Background The Japan Coma Scale (JCS) is the most frequently adopted method for evaluating level of consciousness in Japan. However, no validated method for converting the JCS to the Glasgow Coma Scale (GCS) exists. The aims of the present study were to develop and validate a method to convert the JCS to GCS. Methods This is a multicenter retrospective analysis involving three emergency departments (EDs) in Japan. We included all adult patients who visited the ED between 2017 and 2020. The participating facilities were divided into two cohorts—one cohort to develop a table to convert the JCS to GCS (development cohort), and the other cohort to validate the conversion table (validation cohort). The conversion table of the JCS to GCS was developed based on the median values of the GCS. The outcome was the concordance rate between the JCS and GCS. Results We identified 8,194 eligible patients. The development cohort included 7,373 patients and the validation cohort included 821 patients. In the validation cohort, the absolute and relative concordance rates were 80.3% (95% confidence interval, 77.4–82.9%) and 93.2% (95% confidence interval, 91.2–94.8%), respectively. Conclusion This study developed and validated a novel method for converting the JCS to GCS. Assuming the offset by a single category between the JCS and GCS is acceptable, the concordance rate was over 90% in the general adult patient population visiting the ED. The conversion method may assist researchers to convert JCS scores into GCS scores, which are more commonly recognized among global audiences.


INTRODUCTION 1
An objective assessment for level of consciousness is critical in the emergency department 2 (ED). The Glasgow Coma Scale (GCS), first introduced in 1974, has been widely adopted 3 internationally and has become the de facto gold standard for evaluating conscious level.
[1] 4 On the other hand, the Japan Coma Scale (JCS) is the most commonly applied method for 5 assessing patients' consciousness level in Japan given its simplicity and applicability. [2-4] 6 The JCS consists of four main grades of consciousness based on reactive eye-opening which 7 can be summarized as follows: 0, alert consciousness; 1-3 (single-digit), awake without any 8 stimuli; 10-30 (double-digits), arousable by some stimuli but reverts to previous state if 9 stimulus stops; and 100-300 (triple-digits), unarousable by any stimuli (Table 1). [3] Based 10 on these grades, the JCS has 10 categories of scores ranging from 0 to 300. In contrast, the 11 GCS has 13 categories of scores. 12 The JCS is adopted as a standard prehospital criterion for hospital selection and is 13 widely used by hospital and emergency medical technicians throughout Japan. In addition to 14 this, large-scale Japanese medical databases, including nationwide administrative claims and 15 discharge data, often contain the JCS scores as an indicator of consciousness level instead of 16 GCS. [5][6][7] However, there are no validated methods for converting the JCS to GCS scores. 17 Due to the lack of a validated conversion method, the findings reported in clinical research 18 conducted in Japan are often discounted overseas. Devising a JCS-GCS conversion method is a critical aspect of bridging the divide between these two systems which could lead to the 1 progress and development of research in the field of emergency medicine. 2 The aims of the present study are (i) to demonstrate the relationship between the JCS 3 and GCS, (ii) establish a method for converting the JCS to GCS, and (iii) evaluate the 4 validity of the conversion method using actual cases documented in the general ED setting. 5 6 METHODS 7

Design and settings 8
The present study is a multicenter retrospective cohort analysis involving three EDs in Japan 9 between January 2017 and October 2020. The actual periods of data collection differed 10 between facilities. Participating institutions included Hitachi General Hospital (data 11 collection periods, January 2017 to October 2020), Saiseikai Utsunomiya Hospital (April 12 2020 to September 2020) and the Japanese Red Cross Society Kyoto Daiichi Hospital (April 13 2020 to October 2020). All participating healthcare facilities are tertiary-care emergency 14 hospitals and board specialist training centers certified by the Ministry of Health, Labor, and 15 Welfare in Japan and the Japanese Association for Acute Medicine, respectively. There are 16 approximately 13,000-15,000 ED visits and 5,000-6000 patients transported via emergency 17 medical services annually for each participating facility. [8] 18

Data source and collection 1
In these hospitals, the medical charts are maintained using the NEXT Stage ER system (TXP 2 Medical Co., Ltd. Tokyo, Japan) which supports healthcare professionals record clinical 3 information such as chief complaints, history of present illness, past medical history, 4 medications used, physical assessments, vital signs, and clinical diagnoses as structured data. 5 The clinical diagnoses were recorded with free-text and the information was automatically We identified consecutive data on patients who visited the above EDs between the study 12 period. Patients aged <16 years and those missing either the JCS or GCS on ED arrival were 13 excluded from the present study. 14 15

Variables 16
Patient characteristics included age, sex, emergency medical service use, level of 17 consciousness at the timing of ED arrival, physician's primary diagnosis at the ED and 18 disposition of the ED (admission or discharge). The level of consciousness was recorded 19 using the 10 JCS categories and 13 GCS categories (scores ranging 3-15). The scores were 1 evaluated and recorded by the emergency physicians or nurses upon patient arrival at the ED. 2 The physician's diagnosis at the time of ED discharge was divided into the following 17 3 categories based on the first letter of the ICD-10 code: infectious and parasitic diseases 4 (ICD-10 codes, A and B); neoplasms (C and D); endocrine, nutritional and metabolic diseases 5 (E); mental and behavioral disorders (F); diseases of the nervous system (G); diseases of the 6 eye, adnexa, ear and mastoid process (H); diseases of the circulatory system (I); diseases of 7 the respiratory system (J); diseases of the digestive system (K); diseases of the skin and 8 subcutaneous tissue (L); diseases of the musculoskeletal system and connective tissue (M); 9 diseases of the genitourinary system (N); pregnancy, childbirth, puerperium and perinatal 10 condition (O and P); symptoms, signs and abnormal clinical and laboratory findings, not 11 elsewhere classified (R); injury, poisoning and certain other consequences of external causes 12 (S and T); others (X, Z and Q); and missing data. 13 14

Development and Validation Cohorts 15
The participating facilities were divided into two cohorts: i) development cohort (Hitachi 16 General Hospital) used to develop the table to convert the JCS to GCS, and ii) validation 17 cohort (Saiseikai Utsunomiya Hospital and the Japanese Red Cross Society Kyoto Daiichi 18 Hospital) used to validate and assess the concordance rate. The outcome of the present study 19 was the concordance rate between the JCS and GCS. 1 2

Statistical analysis 3
To describe patient characteristics in the development and validation cohorts, continuous 4 variables were reported as median and interquartile range (IQR) and categorical variables 5 were reported as count and percentage. 6 In the development cohort, we described the median value of the GCS corresponding 7 to each JCS score. Then, we developed a table to convert the JCS to GCS based on the 8 median values. 9 For external validation, the developed conversion table was applied to the validation 10 cohort and the concordance rates based on the table conversions were described. "Absolute 11 concordance" was defined as the percentage of exact concordance with the developed 12 conversion table and "relative concordance" was defined as the percentage of concordance 13 which allows for offset by a single GCS point (+/-1). 14 We calculated the proportion of outcomes and the 95% confidence interval (CI) using the 15 Clopper-Pearson interval. Stata/BE 17 (Stata Corp, College Station, TX, USA) and R version 16 4.1.2 (The R Foundation, Vienna, Austria) were used for statistical analysis. 17

Sensitivity analysis 18
Because the percentage of patients with alert consciousness influence the outcome (i.e. 19 concordance rate), we performed a sensitivity analysis excluding patients with alert 1 consciousness (JCS=0). The outcomes of the sensitivity analysis were the same as the 2 primary analysis (absolute and relative concordance). 3 4 Ethics 5 The study protocol was approved by the Ethics Committee of TXP Medical Co., Ltd 6 (approval number, TXPREC-002). The requirements for informed consent were waved due to 7 the retrospective nature of the study. Only anonymized data was collected from each 8 participating institution. 9 10 RESULTS 11

Study participants 12
After application of the inclusion and exclusion criteria, we identified 8,194 eligible patients 13 during the study period (Figure 1). The development cohort included 7,373 patients from one 14 hospital and the validation cohort included 821 patients from the remaining two hospitals. to convert the JCS to GCS (Table 3). Using this table, if a patient in the validation cohort had 8 a JCS of 30, absolute concordance would be a GCS score of 9. Relative concordance using 9 the same example would be a GCS ranging between 8 and 10. In the validation cohort, 10 absolute concordance was 80.3% (95%CI, 77.4-82.9%) and relative concordance was 93.2% 11 (95%CI, 91.2-94.8%). In the sensitivity analysis excluding patients with alert consciousness 12 (JCS=0), absolute concordance was 46.1% (95%CI, 40.0-52.2%) and relative concordance 13 was 80.5% (95%CI, 75.3-85.1%) ( Table 4). 14 15 16

DISCUSSION 17
In the present study, we established a method to covert the JCS to GCS for use in general ED 18 indications using a multicenter ED database in Japan. Additionally, we evaluated the validity 19 of the conversion method using the concordance rate between the JCS and GCS. Absolute 1 concordance was over 80% and the relative concordance exceeded 90%. Although absolute 2 concordance in patients without alert consciousness was below 50%, the relative concordance 3 rate was preserved at over 80%. 4 The JCS is widely used not only in clinical practice, but also in countless clinical 5 studies in Japan. [5][6][7] Previous studies have demonstrated that the JCS level was beneficial 6 in predicting in-hospital mortality in trauma patients. [3, 4] Another study revealed that the 7 JCS was associated with activity of daily living at 30 days after stroke and mortality in stroke 8 patients.
[2] The JCS is a formidable coma scale; however, the JCS is exclusively utilized in 9 Japan, which is one of the factors impeding the direct application of Japanese clinical study 10 outcomes to overseas practice. The present study is uniquely designed to help translate the 11 JCS into a more broadly adopted consciousness scoring system-the GCS-in order to 12 facilitate dissemination of clinical studies utilizing Japanese databases to a global audience. 13 The sum of the GCS score is used in clinical settings as well as several inherent differences between the components evaluated. To the best of our knowledge, there 1 are no validated methods to convert the JCS to GCS. The present study established and 2 validated a novel method for converting the JCS to GCS. Assuming the offset by a single 3 category between the JCS and GCS is acceptable, the concordance rate was over 90%. The 4 established conversion method in the present study can be broadly utilized by clinical studies 5 using real world big data. 6 One of the strengths of this study is its multi-centered nature, which allowed for 7 enhanced validity. Additionally, we included all emergency patients-not just trauma or 8 stroke-which ensures the broader applicability of the conversion method described in this 9

study. 10
There are several potential limitations that should be acknowledged. not apply a regression analysis in the development of a clinically convenient but highly 19 functional conversion table. Third, the outcomes in the present study (i.e., concordance rate) 1 could have been influenced by the proportion of patients with alert consciousness who may 2 have higher concordance with their respective scores (JCS=0 and GCS=15). To address this 3 issue, we conducted a sensitivity analysis excluding patients with alert consciousness 4 (JCS=0). Fourth, among 66,467 ED visits, 43,749 (66%) patients were excluded due to 5 missing JCS or GCS values. This may have influenced the outcomes and inadvertently 6 introduced a selection bias; however, it can be assumed that level of consciousness was 7 recorded using either the JCS or GCS (and not both) because most patients visiting the ED 8 are generally alert and oriented (e.g., acute respiratory tract infections). Patients who were 9 evaluated using both the JCS and GCS may have presented with unique factors (e.g., when 10 the patient's state of consciousness was difficult to assess via only a single coma scale). 11 Finally, the JCS and GCS should be assessed independently because the components utilized 12 for evaluation differs. The conversion table established in this study may not ensure absolute 13 conversion of the JCS to GCS in individual patients; however, has displayed promising 14 results with regard to the patient population examined in this study. 15 In conclusion, this multicenter cohort study generated a novel method to convert the JCS 16 to GCS and validated concordance between the JCS and GCS. Assuming the offset by a 17 single category between the JCS and GCS is acceptable, the concordance rate was over 90% 18 in the general adult patient population visiting the ED. The conversion method described in 19 the present study may assist Japanese researchers to translate JCS into the GCS format which 1 is more commonly recognized among global audiences.