BACKGROUND
On 11th March 2020, the World Health Organization declared a pandemic caused by the coronavirus with 118.629 identified cases and 4.292 confirmed deaths. Up to date, 252 million cases and 5 million deaths have been identified as caused by COVID-19. An epidemic situation is characterized by an overload of patients suffering a particular clinical condition and needing acute medical attention in a short period. Usually, the pathogen n causing the epidemic is either new or emergent, and the knowledge a priori is limited. Information is crucial for public health authorities to establish policies to prevent transmission. Thus, the cycle of knowledge acquisition must be efficient and as short as possible. An interdisciplinary team adapted the electronic health record alert systems for real-time data tool collection for clinical characterization and epidemiological surveillance. This system has been working from the Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) first outbreak up to date
OBJECTIVE
To share the experience of handling COVID-19 and non-COVID-19 patients' circuits through an Electronic Health Record (EHR) alert system during the pandemic. This system allowed the creation of a COVID-19 hospitalized patient cohort, with implications in the hospital circuit management, patients risk stratification and secondary use for research projects in a period of high uncertainty. Additionally, its integration as an epidemiological surveillance tool favored the submission of updated information to public health authorities.
METHODS
Almost 30,000 alerts related to COVID-19 were activated in the EHR. Overall, the most frequent were “COVID-19 ruled out” (N = 12,438) followed by “COVID-19 Confirmed Case” (N = 8,999). Up to 13,106 patients (65.7%) were labeled with just one alert during their in-patient stay, while 6,857 (34.3%) received two or more labels. For the alert sequences, 96% were considered logical sequences, 3,1% as low-quality logic sequences, and less than 1% aberrant sequences. Although some temporal variations, all periods had a high rate of logical sequences achieving more than 95%. Preventive medicine professionals activated most COVID-19 alerts and acted as auditors for data quality. When possible, automatic alerts were in place, which became the most frequent.
RESULTS
Almost 30,000 alerts related to COVID-19 were activated in the EHR. Overall, the most frequent were “COVID-19 ruled out” (N = 12,438) followed by “COVID-19 Confirmed Case” (N = 8,999). Up to 13,106 patients (65.7%) were labeled with just one alert during their in-patient stay, while 6,857 (34.3%) received two or more labels. For the alert sequences, 96% were considered logical sequences, 3,1% as low-quality logic sequences, and less than 1% aberrant sequences. Although some temporal variations, all periods had a high rate of logical sequences achieving more than 95%. Preventive medicine professionals activated most COVID-19 alerts and acted as auditors for data quality. When possible, automatic alerts were in place, which became the most frequent.
CONCLUSIONS
The EHR integrated system favored in-hospital management of patients during the COVID-19 pandemic. It was helpful for both the institution and the health system, representing an example of interlevel integration. The performance was adequate and robust, with insights at different levels: infection control, patient safety, research, and pandemic response. Preventive Medicine teams should maximize EHR solutions for epidemiological surveillance.
CLINICALTRIAL
Not required.