Screening COVID-19 Patients Using Safe Practice Score System in Non-COVID Radiology Departments.

Background: A prospective study was designed to evaluate the role of safe practice score system (SPSS) for targeted screening of patients in the non-COVID radiology department with the objective to (a) determine the incidence of COVID-19 in patients visiting non-COVID facilities, (b) to determine the sensitivity and specificity with predictive value of SPSS system to identify high-risk COVID patients undergoing ultrasound and echocardiographic examinations, (c) determination of relative risk (RR) estimation for health-care staff and other visiting patients and attendants, and (d) overall impact of the use of SPSS in the prevention of spread of the disease in the society.

Methods: The study comprised a cohort of 688 patients who initially presented as non-COVID patients for ultrasound and echocardiography tests to a non-COVID health-care facility. Patients were divided into low-, moderate-, and high-risk COVID-19 probability groups based on SPSS screening scores of 0-3, 4-6, and more than 6, respectively. All high-risk patients with SPSS of more than 6 were advised real-time polymerase chain reaction (RT-PCR) or plain high-resolution computed tomography chest for the presence of COVID-19 before the diagnostic test could be done and all results were analyzed statistically.

Results: Four hundred and forty-four low-risk patients had a median score of 3, and all underwent their radiological examinations. One hundred and forty patients were in the moderate-risk group and had a median score of 5. Thirty-six patients had a score of 6, out of which 12 patients were upgraded to score 6. One hundred and four patients were in the high-risk category group. Out of these, 20 patients were screened as RT-PCR negative and got their ultrasound examinations. The sensitivity and specificity of SPSS in categorizing patients into low and high risk were 100% and 95%, respectively, with false positive and negative of 4% and 0% with a positive and negative predictive value of 77% and 100%, respectively. The use of SPSS reduced the RR ratio for health-care workers and other staff from 7.9 to 2.6 ( = 0.001). The overall incidence of COVID-19 disease in patients visiting non-COVID hospitals was 12% during this time period. Screening and detection by the use of SPSS had a positive epidemiological impact and saved 33,000 people from getting infected when calculated by the susceptible-exposed-infectious-recovered (SEIR) pandemic model in this period of 90 days.

Conclusion: SPSS categorized patients in low-, moderate-, and high-risk pretest COVID-19 probability categories accurately with good sensitivity and specificity and was useful in preventing the spread of disease with reduced RR to 2.6 for medical staff and other hospital patients and also helped to contain spread in the society.