Current research was analytical, retrospective, cross-sectional which was conducted to evaluate patients with COVID-19 cared at the Regional Hospital of Lambayeque (RHL) in Peru, during the period of COVID-19 health emergency, between both waves: First: March to December 2020 and second: January to September 2021. Lambayeque, a northern region of Peru, is divided into three provinces with a population of 1,197,260 citizens (according to the 2017 Population Census).

We used the clinical follow-up database of the RHL Epidemiology Office to obtain preliminary information of patients with COVID-19. This information was corroborated and incorporated to the epidemiological notification database that was retrieved from the Peruvian Epidemiological Surveillance System (NotiWeb) due to its epidemiological purpose, collects more complete information on the clinical–epidemiological profiles of patients as well as hospitalization and death outcomes. All people with a confirmed diagnosis seen at the RHL were simultaneously reported in the NotiWeb system.

The patients treated for COVID-19 at the RHL from March 2020 to September 2021. RHL was a level III facility according to the Peruvian Ministry of Health (MINSA), meaning that it has been the largest care center in northern Peru and Lambayeque during the COVID-19 pandemic.

The sample included patients with confirmed COVID-19 using routine diagnostic tests (serological/molecular/antigenic). Inclusion of individuals were performed in all patients attended at the RHL, regardless of being new or continuing MINSA users. Patients with no records and incomplete records were eliminated from the study.

The outcomes corresponded to clinical and epidemiological characteristics of COVID-19: signs, symptoms, daily number of cases, and death associated with the disease, which were registered at the time of notification. The exposure was the epidemic wave, which corresponded to the waves: March–December 2020 and January–September 2021. The study variables were divided in 1) epidemiological data: sex, age (continuous data), categorized age, presence of comorbidities (cancer, chronic lung disease, diabetes, obesity, cardiovascular disease, HIV), and 2) COVID-19 clinical data: daily number of cases (continuous data), death associated with the disease (no, yes), cough, odynophagia, chills, fever, respiratory distress, nasal congestion, general malaise, diarrhea, nausea/vomiting, headache, irritability, muscle and abdominal pain, pharyngeal exudate, conjunctival infection, seizure, and dyspnea.

The information from the NotiWeb was compared with the clinical follow-up data by the national identification number as the identification code. Then, a quality review was conducted to detect inconsistent, and incomplete data. Next, a variable called “pandemic wave type” was created considering the clinical and epidemiological profiles according to each wave. It was identified the clinical pattern of each wave.

By Stata v.16.0 the data were analyzed. The calculation of central tendency and dispersion was estimated in the descriptive analysis of numerical variables. Absolute and relative frequencies were calculated. It was conducted the chi-square homogeneity in the bivariate evaluation after evaluation of the expected frequency assumption. To compare the categorical clinical and epidemiological variables between patients attended during both waves was used the Fisher’s exact test. After calculating the normal distribution and homoscedasticity, it was conducted the Student’s t-test or Mann-Whitney U test. The statistical significance was established as p-values of <0.05. To determine the differences in the characteristics of patients with COVID-19 between both waves, a simple regression analysis was performed considering the exposure of each wave. A Poisson statistical model with log link function and robust variance was used to calculate prevalence ratios (PR) (95% CI).

The Ethics Committee of the RHL approved the research protocol, following the ethical guidelines of the Declaration of Helsinki. It was maintained the confidentiality of the patient’s information. Non-identifying keys were used to handle and examine the data. The participants completed a informed consent.

Our results are comparable to similar studies conducted in different settings. We have observed that adults were the main affected group in both waves, which differs from previous reports in Spain showing that young people were mostly affected during the second wave (Iftimie et al. 2021; Mollinedo-Gajate et al. 2021). This is also different from a study in Denmark reporting no differences of severity in both waves, according to age (Buttenschøn et al. 2022). The present study also showed comorbidities were more frequent in the second wave. This result differs from the outcomes reported in India by Kumar et al. (2021) that showed a lower frequency of comorbidities in the second wave (e.g., hypertension, diabetes, and chronic kidney disease). Mortality due to COVID-19 was lower in the second wave, according to our study. The registry of fewer deaths was also observed in Japan (Saito et al. 2021). In France, patients were no different between both waves in similar age, ICU scores and comorbidities (Contou et al. 2021). Results in Australia showed that activation of blocking measures when the second wave cases indicated exponential growth would be effective to reduce COVID-19 cases (Milne et al. 2021). In Italy, the relaxation of individual behaviors, poor and contradictory communication to the population, and the reopening of schools and businesses made the second wave have been worse (Chirico et al. 2021).

The differences observed in our study may be related to characteristics of the population and the epidemic waves in northern Peru. In general, there were more adult patients and a slightly higher frequency of comorbidities. However, vaccination rates were present during the second wave, which could have protected patients from severity and deaths in this region (Valladares-Garrido et al. 2022). This can be reflected in an increasing infection rate due to predominant variants of concern like the Beta and Lambda variants, but without presenting fatal outcomes. Recognizing this pattern may help better understand the behavior of the outbreaks and provide resources efficiently to control the costs to the health care system in northern Peru.

Males were the most affected in both waves, a pattern like the national and international records of patients infected with COVID-19 (Balacchi et al. 2021; Zeiser et al. 2022). This could be explained by the theory that androgens such as testosterone promote SARS-CoV-2 infection through the activation of a transmembrane protease (TMPRSS2) (Baughn et al. 2020; Lukassen et al. 2020; Matsuyama et al. 2020; Ragia and Manolopoulos 2020; Wambier and Goren 2020; Abbasi et al. 2021).

The average age of patients was slightly lower in the second wave. This difference has also been found in other countries such as India (Kumar et al. 2021), Germany (Graichen 2021), Japan (Saito et al. 2021), South Korea (Seong et al. 2021), and Iran (Jalali et al. 2020). In Peru, the predominant variants during the second wave were Lambda (C37 or the Andean variant) and Beta (B.1.351) (Aguilar-Gamboa et al. 2021) due to their greater transmissibility, which displaced other variants, specifically Gamma (P.1) that was concentrated in Lambayeque (Ministerio de Salud 2021a). It has been observed that the Gamma variant affects a greater proportion of people aged >65 years, whereas the Lambda variant affects more children and adolescents (Instituto de Salud Pública de Chile 2021). Restriction measures implemented during the first wave became less strict during the second wave, which resulted in younger people resuming their work activities (Hoogenboom et al. 2021), and even without having received immunization against COVID-19, they were more exposed to SARS-CoV-2 infection. Furthermore, early-care-seeking among younger people may have resulted in greater access to health services, leading to a relative increase of cases in young people during the second wave compared with the first wave (Kumar et al. 2021). Another explanation may be that older adults, being more vulnerable, died during the first wave, generating a shift to a younger group during the second wave (Hoogenboom et al. 2021).

In the second wave, the prevalence of COVID-19 infection in children was 250% higher than that in the first wave (PR: 3.50; 95% CI: 2.79–4.40), whereas the prevalence in older adults decreased by 17% (PR: 0.83; 95% CI: 0.77–0.90) compared to that in the first wave. These findings could be due to the start of vaccination in older adults (as of April 26, 2021), which protected against COVID-19 in second wave (that lasted until the end of December 2021 approximately), and consequently, the pediatric population remaining unvaccinated was affected in a greater proportion.

In second wave, the incidence of respiratory signs and symptoms such as cough (PR: 1.20; 95% CI: 1.15–1.24), dyspnea (PR: 1.43; 95% CI: 1.33–1.54), nasal congestion (PR: 2.17; 95% CI: 1.90–2.49), and abnormal pulmonary auscultation (PR: 2.73, 95% CI: 2.32–3.21) was significantly increased compared to that in the first wave. This finding is different from an Italian study in which during the first wave, the proportion of cases with dyspnea (71.7% vs. 37.1%, OR: 4.29) and cough (45% vs. 22.8%, OR: 2.76) was higher in the second wave (Bongiovanni et al. 2021). This could be due to better recording of information by the epidemiological surveillance system of the epidemiological surveillance center of MINSA, better training and awareness for data capture by frontline staff, and lower patient demand, which facilitated data recording by the health personnel. Compared with the beginning of the pandemic, in which cough, respiratory distress, malaise, odynophagia, and diarrhea were the most frequent symptoms of SARS-CoV-2 positivity, we recognized that during the second wave, positivity was more frequent in males, and cough, odynophagia, and malaise were significantly lower compared with the first wave. Respiratory distress and diarrhea were frequent in both waves of the pandemic (Vera-Ponce et al. 2021).

Current research shows that during the second wave, there was a significant increase in the number of patients with comorbidities in general. This is a different finding from an Indian study that found that patients examined in the second wave had a lower percentage of comorbidities (45.9% in the second wave vs. 55.9% in the first wave) (Kumar et al. 2021). The proportion of cases with cardiovascular illness (12.5% vs. 16.7%; PR: 1.34), diabetes mellitus (9.3% vs. 11.5%; PR: 1.23), obesity (0.7% vs. 8.1%; PR: 11.27), and chronic kidney disease (2.8% vs. 6.7%; PR: 2.40) was increased during the second wave. This finding coincides with a multicenter study from Spain, Andorra, and Ireland, which reported that patients in ICU for COVID-19 during the second wave had a higher proportion of comorbidities such as diabetes mellitus (20.9% vs. 26.7%; p < 0.001), obesity (32.2% vs. 42.3%; p < 0.001), and chronic kidney disease (4.6% vs. 6.8%; p = 0.005) (Carbonell et al. 2021).

Another study in Mexico during the second wave reported 5.4% of cases of chronic kidney disease (Puicón-Suárez et al. 2022). However, this is different from a study in India that reported that the proportion of patients with cardiovascular disease (6.9% vs. 4.2%; p < 0.001), diabetes mellitus (26.5% vs. 21.5%; p < 0.001), and chronic kidney disease (3.9% vs. 3.3%; p = 0.03) decreased during the second wave (Kumar et al. 2021). This could be because prevention measures were stricter in the first wave so that patients with comorbidities could have greater prevention during the first wave than during the second wave. It could also be a consequence of better identification of patients with comorbidities due to the implementation of citizen surveillance strategies of high-risk populations and/or people with comorbidities, which was established in the plan for a potential second wave by COVID-19 in the Ministry of Health of Peru in December 2020 (Ministerio de Salud 2021b).

In this study, the proportion of cases with HIV (0.1% vs. 0.4%; PR: 5.43), cancer (1.2% vs. 3.9%; PR: 3.32), and pregnancy (2.9% vs. 4.7%; PR: 1.61) increased between waves. One Indian study corroborates the increase in the proportion of cancer cases (1.8% vs. 1.9%; p = 0.5), although not significant (Kumar et al. 2021). However, another Indian study did find a significant increase (2% vs. 5.9%; p < 0.001) (Kumar et al. 2021). Similarly, a multicenter study conducted in Europe identified that in the second wave, the proportion of cases with overall immunosuppression increased (4.2% vs. 8.3%; p < 0.001) compared to that in the first wave (Carbonell et al. 2021).

Genomic surveillance is limited in developing countries; therefore, the findings of this study could be useful for epidemiological follow-up and for understanding the differences between both waves. Continuous epidemiological surveillance could help to be prepared for a possible fourth wave concerning the management of COVID-19 and prioritizing treatment for the most vulnerable population.

Current research has important limitations. Initially, a possible measurement bias existed during data collection due to different registration personnel. Moreover, our results reflect the profile of patients in the first and second waves from a single hospital in Peru. Hence, there could be a selection bias, where these findings do not represent the characteristics of patients with COVID-19 in the entire country, and there could also be significant differences concerning other regions of the country. Additionally, the comparison of clinical and epidemiological characteristics lacks detailed information on variants of concern. In future studies, a robust statistical difference may be achieved adjusting the regression analysis by relevant factors like age and sex. Despite these limitations, the strength of this study is that the sample size is large and obtained from a reference hospital for COVID-19 care, meaning that data have been captured from patients residing in not only Chiclayo but also the entire northern Peruvian macro region. This allows an adequate comparison and description of the pandemic waves.