Translation and validation of the Korean version of acute cystitis symptom score

Urinary tract infection (UTI) is the most common bacterial infection in general practice [1, 2, 3]. Almost half of all women have at least one episode of UTI during their lifetime [4]. Acute uncomplicated cystitis (AUC) accounts for the greatest number of UTIs, particularly in young sexually active women. AUC is usually diagnosed based on clinical symptoms and laboratory findings including pyuria or a positive urine culture in primary health care.

However, depending on types of hospitals, laboratory evaluation could not be performed. Thus, physicians are asked to prescribe empirical antibiotics based on only complaints of patients in the absence of urine culture and sensitivity data. Fortunately, the probability of AUC in a woman with dysuria, urinary frequency, or hematuria is reported to be about 50%. Specific combinations of symptoms raise the probability to be more than 90% [5]. Therefore, questionnaires and symptom scores can be used to assess the diagnosis and severity of AUC more accurately.

Acute Cystitis Symptom Score (ACSS) is a simple self-reporting questionnaire initially developed in Uzbek language to help diagnose AUC in primary healthcare settings [6]. ACSS questionnaire consists of an 18-item self-reporting questionnaire (6 questions about typical symptoms of AUC, 4 questions regarding differential diagnoses, 3 questions on quality of life [QoL], and 5 questions on additional conditions that might affect therapy). It was further translated into Russian [7], German [8], Hungarian [9], Italian [10], American English version [11], and so on (http://www.acss.world) and validated. Therefore, we aimed to develop a Korean version of the ACSS and validate it using Korean-speaking women.

The diagnosis of AUC is usually based on clinical symptoms and laboratory findings including pyuria or a positive urine culture in primary health care. However, as recommended in various guidelines, an accurate diagnosis of AUC needs to focus on a history of lower urinary tract symptoms (dysuria, frequency and urgency) and the absence of vaginal discharge or irritation [15]. In addition, depending on types of hospitals, laboratory evaluation could not be performed. Physicians in primary care clinics are often asked to prescribe empirical antibiotics only based on complaints of patients in the absence of urine culture and sensitivity data. Under such health care environments, questionnaires and symptom scores can be used to perform the correct diagnosis of AUC.

Previously, the Urinary Tract Infection Symptoms Assessment (UTISA) questionnaire from Bayer Health Care Pharmaceuticals Global Health Economics and Outcome Research, a 14-item self-reporting questionnaire about the severity and bothersomeness of seven key uncomplicated UTI symptoms [16], has been translated into Korean with linguistic validation. However, statistical information such as predictive and discriminative ability were not evaluated [17]. Thus, it is still not widely used in practice or clinical research in Korea. On the other hand, the ACSS has been translated and validated into various languages.

The ACSS questionnaire was initially developed in Russian and Uzbek language [6]. In a combined analysis for both versions, Cronbach’s alpha for reliability of the ACSS was 0.89. Sensitivity and specificity to predict AUC were 94% and 90%, respectively, at cut-off score of 6 for ‘Typical’ domain. Cronbach’s alpha for the Russian version of the questionnaire was 0.86 and the area under the curve in the ROC analysis was 0.96. Symptom scores were positively correlated with laboratory results [7]. In European countries, Cronbach’s alpha for the German ACSS was 0.87. A cut-off score of 6 for the ‘Typical’ domain significantly predicted AUC with a sensitivity of 94.7% and a specificity of 82.4% [8]. The Hungarian version also showed high levels of reliability and validity with sensitivity and specificity of 90% and 97%, respectively, at cut-off score of 6 for the ‘Typical’ domain [9]. The Italian version of the ACSS reported a sensitivity of 92.5% and a specificity of 97.8% at a cut-off score of 6 for the domain of typical symptoms [10]. Recently, the American-English ACSS showed excellent results of psychometric parameters and diagnostic values in clinical validation [11]. Our results are also consistent with validation results of ACSS in other languages. The Korean version of the ACSS showed high levels of reliability and predictive ability. And the sum score of typical symptoms show the highest balance for diagnostic sensitivity and specificity. However, the individual typical symptoms such as urinary urgency, pain in lower abdomen, and hematuria showed low levels of reliability and predictive ability (area under the ROC curve=0.744, 0.725, and 0.651, respectively).

The shortcomings of the study are (1) performed in a single center, and (2) in a limited number of patients (n=31) and controls (n=19). For all typical symptoms and QoL items, however, a significant difference could be found between patients and controls. Nevertheless, confirmation of the results in a larger multicenter study would be desirable. Finally, the original ACSS was composed of a diagnostic part (part A) and a follow-up part (part B). The follow-up part included a domain of ‘dynamics’ with five questions about changes of symptoms in addition to the same questions as those in the diagnostic part [18]. Therefore, further studies are needed to investigate applicability of the ACSS as a practical questionnaire for patient-reported outcome assessment in Korea.

In this study, a Korean version of the ACSS was tested in 50 women (31 AUC patients and 19 controls). It showed high levels of reliability and validity, similar to other validated versions of the ACSS in different languages. Thus, it will play an important role in practice and/or clinical research for the diagnosis and treatment efficacy monitoring of Korean-speaking women suffering from AUC.