Skip to main content

Comparing the first and second waves of COVID-19 in a tertiary university hospital in Barcelona

Cite this dataset

Cossio Gil, Yolima et al. (2021). Comparing the first and second waves of COVID-19 in a tertiary university hospital in Barcelona [Dataset]. Dryad.



To compare the flow of COVID-19 patients in emergency rooms and hospital wards between the pandemic; first and second waves; at the University Hospital of Vall d’Hebron (Barcelona, Spain), and to compare the profiles, severity and mortality of COVID-19 patients between the two waves.


Retrospective observational analysis of COVID-19 patients attending the hospital from February 24 to April 26, 2020 (first wave) and from July 24, 2020, to May 18, 2021 (second wave). We analysed the data of the Electronic Medical Records on patient demographics, comorbidity, severity and mortality.


The daily number of COVID-19 patients entering the ER dropped by 65% during the second wave compared to the first wave. During the second wave, patients entering the ER were significantly younger (61 vs 63 y.o. p<0.001) and less severely affected (39% vs 48% with a triage level of resuscitation or emergency; p<0.001). ER mortality declined during the second wave (1% vs 2%; p<0.000). The daily number of hospitalised COVID-19 patients dropped by 75% during the second wave. Those hospitalised during the second wave were more severely affected (20% vs. 10%; p<0.001) and were derived to the intensive care unit (ICU) more frequently (21% vs 15%; p<0.001). Inpatient mortality showed no significant difference between the two waves.


Changes in the flow, severity and mortality of COVID-19 patients entering this tertiary hospital during the two waves may reflect a better adaptation of the health system and the improvement of knowledge on the disease.



The study included 8,684 distinct COVID-19 patients who attended the ER and/or were hospitalised. Cases were identified using the ICD-10-CM diagnostic codes recorded in the hospital-discharge data: B34.2 and B97.29 (from February 2020) and ICD-10 code U07.1 (from July 1, 2020). Discharge diagnoses of hospitalised patients were registered by specialised coders. To guarantee that we were analysing and comparing data from patients affected by the COVID-19 disease, we only included: i) those patients whose primary diagnosis was COVID-19 or, ii) patients with a COVID-19 secondary diagnosis following a first diagnosis of a respiratory system disease (ICD-10-CM code starting by J) or was described with a Diagnosis Related Groups (DRG) 137 (Infections in major lung inflammations) or 139 (other pneumonia). The selection criteria for the patients attending the ER was to have a confirmed COVID-19 diagnosis, regardless of whether it was a primary or secondary diagnosis.

Variables and data sources

To compare the inpatient and emergency flow activity of the HUVH between waves, we created an indicator measuring the total daily number of patients entering the hospital. To compare the characteristics of COVID-19 patients who entered the hospital emergency department and/or were hospitalized, we used data on patient demographics (sex and age) and comorbidity (The Adjusted Morbidity Groups (GMA)). GMA is a validated morbidity measurement developed and adapted to the Spanish Healthcare System that classifies the population into seven morbidity groups, taking into account the typology of their diseases (acute, chronic, or oncological), and in the case of chronic disease, identifying if it is a single or a multimorbidity. Two additional groups identify pregnant/childbirth women and populations without previous pathologies (18,19).

To compare the severity of patients attending the ER, we included the type of emergency as defined by the Structured Triage System for emergency services implemented in Catalonia, which is based on the Australian-Canadian triage systems, and divides the emergencies into five groups: resuscitation, emergent, urgent, less urgent and non-urgent. Severity of hospitalised patients was analysed according to the following variables: first, the severity of the episode, as measured by the weight of the diagnosis-related groups (APR- DRG, version 36) and its severity. Changes in the DGR weights of COVID-19 related diagnosis occurred in July 2020 as a consequence of starting using the U07.1 diagnosis code for COVID-19. Second, the number of patients accessing the intensive care unit and third, the overall length of stay (LOS) during the hospitalization and the ICU-specific LOS. Finally, to compare the mortality rates between the two waves, we included three variables: overall mortality during the hospitalization, 3-day mortality and mortality of patients who entered the ICU.

Data was obtained from the electronic medical records and extracted using structured query language (SQL), to create an anonymized database.  The data collection period was defined as per the hospital admission date and only discharged patients’ data was included.