Clinical-epidemiologic variation in patients treated in the first and second wave of COVID-19 in Lambayeque, Peru: A cluster analysis.

OBJECTIVES: To identify differences in the clinical and epidemiologic characteristics of patients during the first and second waves of the COVID-19 pandemic at the EsSalud Lambayeque health care network, Peru.
METHODS: An analytical cross-sectional study of 53,912 patients enrolled during the first and second waves of COVID-19 was conducted. Cluster analysis based on clustering large applications (CLARA) was applied to clinical-epidemiologic data presented at the time of care. The two pandemic waves were compared using clinical-epidemiologic data from epidemiologic surveillance.
RESULTS: Cluster analysis identified four COVID-19 groups with a characteristic pattern. Cluster 1 included the largest number of participants in both waves, and the participants were predominantly female. Cluster 2 included patients with gastrointestinal, respiratory, and systemic symptoms. Cluster 3 was the "severe" cluster, characterized by older adults and patients with dyspnea or comorbidities (cardiovascular, diabetes, obesity). Cluster 4 included asymptomatic, pregnant, and less severe patients. We found differences in all clinical-epidemiologic characteristics according to the cluster to which they belonged.
CONCLUSION: Using cluster analysis, we identified characteristic patterns in each group. Respiratory, gastrointestinal, dyspnea, anosmia, and ageusia symptoms were higher in the second COVID-19 wave than the first COVID-19 wave.

Introduction

The coronavirus disease 2019 (COVID-19), caused by Severe Acute Respiratory Syndrome Severe Coronavirus 2 (SARS-CoV-2), began in late 2019 in Wuhan City, China, and spread rapidly worldwide (Guo et al., 2020). Up to March 2022, the World Health Organization (WHO) reported approximately 483 million confirmed cases and around 6 million deaths due to this disease (WHO Coronavirus (COVID-19) Dashboard, n.d.). Likewise, Peru, one of the countries most affected by COVID-19, reported nearly 3 million confirmed cases and 212,157 deaths (Peru, n.d.), reporting the highest mortality rate (652.58 deaths per 100,000 population) worldwide (Mortality Analyses, n.d.).

In Peru, the first wave of COVID-19 began in March 2020, peaking in August and ending in December 2020 (Peru, n.d.), while the second wave began in January 2021, with the maximum peak of infections and deaths in April and culminating in June 2021 (Peru, n.d.). Lambayeque was one of the regions most affected by COVID-19 (Díaz-Vélez et al., 2021a, 2021c), ranking in the highest quartile of mortality and surpassing even the national death rate, with a regional mortality rate of 612.2 deaths (Situación del COVID-19 en el Perú, n.d.). The first wave of COVID-19 began in Lambayeque in March 2020, reached its peak in May, and ended in December 2020, while the second wave started in January and ended in June 2021 (Situación del COVID-19 en el Perú, n.d.).

The clinical presentation of COVID-19 includes cough, nasal congestion, fever, chest pain, and shortness of breath (Griffin et al., 2021; Jiang et al., 2020), varying according to the level of clinical classification (mild, moderate, severe, and critical) (Instituto de Evaluación de Tecnologías en Salud e Investigación, 2021). Previous studies have described clinical differences between the first and second waves, particularly in age groups, symptomatology, and disease severity (Area et al., 2021; Friston et al., 2021; Iftimie et al., 2021; Mocanu et al., 2021; Mollinedo-Gajate et al., 2021; Salyer et al., 2021; Soriano et al., 2021; Vinceti et al., 2021).

In Peru, the second wave presented a more significant acceleration of infections, deaths, and bed occupancy in intensive care units and intermediate care units, affecting mainly the elderly and pregnant women (Huatuco-Hernández et al., 2021). The appearance of the second wave of COVID-19 may have been associated with the circulation of new variants with higher virulence, transmissibility, and lethality, such as the British variant B.1.1.7 and Brazilian variant P.1, reported by the National Institute of Health after performing genomic sequencing (CDC-Perú)., n.d.).

There is no documented evidence on the clinical-epidemiological profile among patients treated in the two waves of SARS-CoV-2 infection in Peru, much less at the regional level. In addition, there is little genomic surveillance at the local level to identify the sequence of the viral genome in order to identify clinical patterns according to the main SARS-CoV-2 strains, and thus, provide an approximation of the clinical-epidemiological differences of both waves, which could represent an effective tool in public health decision-making for the prevention and control of COVID-19 in the Lambayeque region. This study aimed to describe the main clinical and epidemiological differences between the first and second waves of COVID-19 in the Lambayeque region, Peru, by analyzing the clinical follow-up records of patients treated in hospitals of the EsSalud Lambayeque health care network. (Díaz-Vélez et al., 2021b).

Methods

Study design

An analytical cross-sectional study was carried out to determine the clinical and epidemiological differences among patients treated for COVID-19 in institutions belonging to the EsSalud health network in Lambayeque, during the first and second waves of the COVID-19 health emergency.

Population and sample

The population consisted of patients treated for COVID-19 in institutions belonging to EsSalud from March 2020 to September 2021. The sample was made up of 53,912 patients with suspected COVID-19 recorded in the Notification System of the Ministry of Health (NotiWeb-MINSA). Confirmed or suspected COVID-19 patients, new or continuing EsSalud user patients, and individuals who had been treated and notified in the NotiWeb System in EsSalud Lambayeque service networks during the aforementioned period were included. Patients with incomplete clinical records and those with absent clinical records for the variables of interest were excluded.

Instruments and variables

The study variables were collected and analyzed and divided into three sections: general data of the notification and of the patient, including the date of notification, and the classification of the case. General epidemiological variables included: sex, age, categorized age. Clinical variables were subclassified into: a) clinical manifestations, with which the case type was recorded according to severity, date of symptom onset if the patient was isolated, and when they were isolated; b) symptoms the presentation of cough, sore throat, nasal congestion, shortness of breath, fever, chills, malaise, diarrhea, nausea/vomiting, headache, anosmia, ageusia, ear pain, irritability/confusion, muscle pain, chest pain, abdominal pain and joint pain, c) recording of clinical signs, body temperature, presence of pharyngeal exudate, conjunctival injection, seizure, coma, dyspnea/tachypnea; and lastly, d) comorbidities including pregnancy, postpartum period (<6 weeks), presence of cardiovascular disease, diabetes, liver disease, neurological/neuromuscular disease, obesity, immunodeficiency, kidney disease, liver damage, chronic lung disease, and cancer.

Procedures and techniques

Two databases were used in this study: 1) the clinical follow-up database of the "Oficina de Inteligencia Sanitaria de Red Prestacional Lambayeque-EsSalud" and 2) the epidemiological notification file database of the NotiWeb Epidemiological Surveillance System of the National Center for Epidemiology, Prevention and Control of Diseases, Peru.

The NotiWeb database was exported, and was matched with the clinical follow-up database using an identity document as the identifier code. A quality control process was then carried out to identify inconsistent data, and ​​out of range and/or incomplete values. Subsequently, a variable called "type of pandemic wave" was created, in which the clinical-epidemiological characteristics of the patients were grouped according to the period of each pandemic wave. The first wave was defined between the months of March-2020 to December-2020 and the second wave between January-2021 to September-2021. This made it possible to identify the clinical pattern of each wave and identify the clinical-epidemiological differences between each time period. Finally, the data were statistically analyzed.

Statistical analysis

The data were analyzed using the statistical program Stata v16.0 (StataCorp LP, College Station, TX, USA).

In the descriptive analysis of numerical variables, the best measure of central tendency and dispersion was calculated. In the case of categorical variables, the absolute and relative frequencies were estimated. In the bivariate analysis, the chi-square test was used after evaluating the assumption of expected frequencies. The Fisher's exact test (categorical variables) was also used to compare the categorical clinical-epidemiological variables between patients treated in the first and second waves. In the case of numerical variables (age, length of hospitalization), the Student's t-test was used after evaluating the assumption of normal distribution and homoscedasticity. The Mann-Whitney U test was also used. P values ​​less than 0.05 were considered statistically significant.

Additionally, the epidemiological curve was constructed according to the date of onset of symptoms of the first and second epidemic waves. The database was collapsed according to the date of onset of symptoms, and the variable weekly moving average of cases was created. Subsequently, the case trends of each pandemic wave were estimated, plotting the cumulative, daily curve, and weekly moving average.

Cluster analysis: Clustering Large Applications (CLARA) was used as a clustering method. This method is an option for large data since it works with a resampling scheme by taking smaller random subsamples and polling them to propose the results. The silhouette method was used to determine the optimal number of clusters.

To identify groups of patients with similar clinical characteristics we used the clustering algorithm CLARA as implemented by the function clara of “cluster” package in R version 4.1.2. This method is an option for large data since it works with a resampling scheme by taking smaller random subsamples and polling them to propose the results (Kassambara, 2017). We determined the optimal number of cluster based on the silhouette method and epidemiological considerations.

Ethical aspects

The research protocol was approved by the Ethics Committee of the Almanzor Aguinaga Asenjo National Hospital with registration code No. CIE-RPL 047-SET-2021. Additionally, the research protocol was registered in the repository of health research projects of the National Institute of Health-Peru (PRISA-INS). The study was carried out following the ethical principles of the Declaration of Helsinki. The confidentiality of the clinical records of the patients selected for the investigation was preserved. Anonymized codes were used to manage and analyze the information, and only the study investigators had access to the data.

Results

Between March 2020 and September 2021, two heterogeneous waves of positive COVID-19 cases were recorded. The maximum peak of the first wave was observed in August 2020, followed by a decrease in the number of confirmed cases. In December 2020, phase IV of the economic reactivation began, leading to the onset of a second pandemic wave in January 2021, the peak of which was reached in April 2021. The peak of the second pandemic wave was slightly lower than that recorded in the first wave. The epidemiological curve of confirmed cases in the Lambayeque health care network continues to decrease. The weekly moving average is much lower than that observed during the maximum peak of the first and second pandemic waves (Figure 1 ).

Figure 1

Epidemiological curve of confirmed COVID-19 cases in the Lambayeque health care network during the first and second waves according to the date of symptom onset.

Table 1 shows the epidemiological and clinical characteristics of patients seen during the first and second waves of the COVID-19 pandemic in the Lambayeque health care network. Compared to the first wave, the second wave was characterized by patients with a higher mean age (47.92 vs. 44.87), a higher proportion of older adults (31.2% vs. 25.8%) and males (50.3% vs. 43.7%). Regarding signs and symptoms, a higher frequency of cough (56.3 vs. 44.9%), nasal congestion (26.4% vs 18.5%), respiratory distress (29.5% vs. 25.3%), diarrhea (20.9% vs. 17.3%) and headache (39.0% vs. 36.7%), dyspnea (5.7% vs. 3.6%) was observed during the second wave compared to the first, respectively. In addition, compared to the first wave, in the second wave there was a higher proportion of cardiovascular disease (14.4% vs. 11.0%), diabetes (7.4% vs. 6.1%), cancer (1.7% vs. 1.1%) and obesity (5.4% vs. 0.6%), respectively. In contrast, there was a higher frequency of lung disease in the first compared to the second wave (1.0% vs. 0.7%, respectively).

Table 1

Clinical-epidemiological characteristics of patients with COVID-19 during the first and second waves.

Variables	Pandemic wave
	First wave (n=36938)		Second wave (n=16974)
	n	%	n	%
Age (years)*		44.87 ± 20.5	47.92 ± 20.8
Age (categorized)
Child (0-11)	1859	5	626	3.7
Adolescent (12-17)	1451	3.9	535	3.2
Young (18-29)	5570	15.1	2431	14.3
Adult (30-59)	18539	50.2	8092	47.7
Older adult (60-)	9519	25.8	5290	31.2
Gender
Female	20803	56.3	8444	49.8
Male	16135	43.7	8530	50.3.
CLINICAL CHARACTERISTICS
General malaise	16721	45.3	12044	71.0
Cough	16578	44.9	9556	56.3
Sore throat	14003	37.9	7205	42.5
Fever	12162	32.9	6215	36.6
Headache	11640	31.5	6219	36.6
Nasal congestion	6817	18.5	4473	26.4
Muscle pain	6185	16.7	3826	22.5
Respiratory distress	6569	17.8	3586	21.1
Diarrhea	5089	13.8	3096	18.2
Chills	521	1.4	1787	10.5
Chest pain	3923	10.6	1738	10.2
Nausea	2274	6.2	1257	7.4
Anosmia	334	0.9	1158	6.8
Ageusia	290	0.8	977	5.8
Dyspnea	1314	3.6	972	5.7
Abdominal pain	1082	2.9	596	3.5
Ear pain	48	0.1	204	1.2
Pharyngeal exudate	343	0.9	104	0.6
Irritability	298	0.8	79	0.5
Conjunctival injection	108	0.3	44	0.3
Seizure	8	0.0	2	0.0
COMORBIDITIES
Cardiovascular disease	4078	11.0	2445	14.4
Diabetes	2236	6.1	1251	7.4
HIV	42	0.1	12	0.1
Kidney disease	364	1.0	215	1.3
Lung disease	371	1.0	110	0.7
Cancer	399	1.1	281	1.7
Obesity	208	0.6	921	5.4
Pregnancy	1208	3.3	574	3.4

Mean and standard deviation; HIV: human immunodeficiency virus

Cluster analysis

From a purely statistical perspective, four clusters were determined to be the optimal number (Supplementary figure 1). The population was categorized into these four clusters, and comparative tables and figures were made to characterize the patterns. The characteristics of the participants of the four clusters are shown in Table 2 . All the clinical-epidemiological characteristics differed among the groups (p-values<0.005).

Table 2

Characteristics of the participants according to cluster.

Characteristic	OverallN = 53,9121	1N = 16,8321	2N = 14,2841	3N = 11,6801	4N = 11,1161	p-value2
COVID-19 pandemic wave						<0.001
First wave	36,938 (68.5%)	10,700 (63.6%)	8,805 (61.6%)	6,816 (58.4%)	10,617 (95.5%)
Second wave	16,974 (31.5%)	6,132 (36.4%)	5,479 (38.4%)	4,864 (41.6%)	499 (4.5%)
Year						<0.001
2020	36,938 (68.5%)	10,700 (63.6%)	8,805 (61.6%)	6,816 (58.4%)	10,617 (95.5%)
2021	16,974 (31.5%)	6,132 (36.4%)	5,479 (38.4%)	4,864 (41.6%)	499 (4.5%)
Gender						<0.001
Female	29,247 (54.2%)	11,311 (67.2%)	7,775 (54.4%)	3,436 (29.4%)	6,725 (60.5%)
Male	24,665 (45.8%)	5,521 (32.8%)	6,509 (45.6%)	8,244 (70.6%)	4,391 (39.5%)
Age (years)						<0.001
Mean (SD)	45.8 (20.6)	38.9 (18.7)	42.7 (18.1)	63.6 (16.3)	41.7 (20.0)
Median (IQR)	45.0 (31.0, 61.0)	38.0 (26.0, 52.0)	42.0 (29.0, 56.0)	65.0 (53.0, 76.0)	39.0 (28.0, 56.0)
Range	0.0, 103.0	0.0, 100.0	0.0, 103.0	0.3, 103.0	0.0, 101.0
Age group						<0.001
0-4	1,098 (2.0%)	521 (3.1%)	129 (0.9%)	11 (0.1%)	437 (3.9%)
5-9	927 (1.7%)	540 (3.2%)	224 (1.6%)	9 (0.1%)	154 (1.4%)
10-14	1,278 (2.4%)	671 (4.0%)	391 (2.7%)	10 (0.1%)	206 (1.9%)
15-17	1,168 (2.2%)	519 (3.1%)	359 (2.5%)	30 (0.3%)	260 (2.3%)
18-29	8,001 (14.8%)	3,132 (18.6%)	2,549 (17.8%)	238 (2.0%)	2,082 (18.7%)
30-59	26,631 (49.4%)	8,865 (52.7%)	7,891 (55.2%)	4,144 (35.5%)	5,731 (51.6%)
60-79	11,747 (21.8%)	2,347 (13.9%)	2,405 (16.8%)	5,154 (44.1%)	1,841 (16.6%)
80-103	3,062 (5.7%)	237 (1.4%)	336 (2.4%)	2,084 (17.8%)	405 (3.6%)
Case type						<0.001
Asymptomatic	11,157 (20.7%)	278 (1.7%)	159 (1.1%)	122 (1.0%)	10,598 (95.3%)
Symptomatic	42,755 (79.3%)	16,554 (98.3%)	14,125 (98.9%)	11,558 (99.0%)	518 (4.7%)

n (%); SD: standard deviation; IQR: interquartile range

Pearson's Chi-squared test; Kruskal-Wallis's rank sum test; Fisher's exact test

Figure 2A shows that cluster 1 was the most frequent cluster (N=16,832; 31.2%), while the least frequent was cluster 4 (N=11,116; 20.6%). Figure 2B shows the distribution of clusters according to pandemic wave. Cluster 1 predominated in the first (N=10,700; 29%) and second waves (N=6,132; 36.1%). Cluster 4 was the second most frequent in the first wave (N=10,617; 28.7%) and was the least frequent in the second wave (N=499; 2.9%). The distribution of persons by cluster according to epidemiological weeks (EW) is shown in figure 2C. In EW 12-2020, clusters 1 (0.4%) and 2 (0.6%) were the most frequent. Then, a progressive increase of cluster 4 was observed between EW 13 to 25, being predominant between EW 26 to SE 44 and finally, becoming less frequent between EW 45 to 53 of 2020. Additionally, cluster 3 increased and became more frequent in the second wave of 2021, while clusters 1 and 2 fluctuated, but seemed to be the most stable over time.

Figure 2

Cluster analysis according to pandemic wave and epidemiological week (EW).

Figure 3A shows the age distribution according to cluster, with cluster 3 having a greater age distribution. Figure 3B presents the distribution of the age groups. The age group between 30 and 59 years of age was the most predominant in clusters 1 (52.7%), 2 (55.2%) and 4 (51.6%), while in cluster 3, the age group comprising patients between 60 and 79 years of age predominated (44.1%). Figure 3C shows that the female sex was more frequent in clusters 1 (67.2%), 2 (54.4%) and 4 (60.5%), while male sex predominated in cluster 3 (70.6%). In Figure 3D clusters 1 (98.4%), 2 (98.9%) and 3 (99%) were predominantly symptomatic, while cluster 4 was characterized by asymptomatic patients (95.3%).

Figure 3

Clinical characteristics of the participants according to cluster of affiliation (i)

Patients in cluster 3 presented the most severe clinical characteristic. Dyspnea was more predominant in cluster 3 (60.6%) compared to the other clusters, and chest pain was more frequent in cluster 2 (23.6%). The frequency of patients with cardiovascular disease, diabetes and obesity was higher in cluster 3 compared to the remaining groups (Figure 4 ).

Figure 4

Clinical characteristics of the participants according to cluster of affiliation (ii).

Discussion

Clinical-epidemiological variation in the first and second wave of the pandemic

In the second wave of the COVID-19 pandemic, there was a greater number of older adult patients and a lower frequency of children and adolescents. These results are similar to those reported by Soriano et al. in a medical center in Madrid, Spain, in which a slightly higher, albeit not statistically significant, mean age was observed in the second compared to the first wave (Soriano et al., 2021). However, these findings differ from those reported in similar studies conducted in China (Fan et al., 2021), Japan (Saito et al., 2021) and Italy (Vinceti et al., 2021). In a hospital in Spain, younger patients predominated in the second wave (Iftimie et al., 2021). In Sudan, fewer elderly patients with COVID-19 were reported in the third wave compared to the second wave (23.9% vs. 76.1%, respectively), although there were no significant differences (Abd El-Raheem et al., 2021).

There were more patients with chronic diseases in the second than in the first wave. This finding is similar to that reported in Spain, in which a higher frequency of patients with cardiovascular diseases was observed, while the number of diabetic patients decreased, although with no significant differences being found in the analysis (Iftimie et al., 2021). However, our results differ from those reported in Japan, in which a lower prevalence of patients with cardiovascular, cerebrovascular, respiratory, renal, diabetic and obese comorbidities was observed in the second wave compared to the first wave (Saito et al., 2021). Our findings are also contrary to those in Sudan, in which the frequency of patients with comorbidities was markedly reduced in the third wave compared to the second wave of the COVID-19 pandemic (Abd El-Raheem et al., 2021).

Cluster analysis based on CLARA

The 4 clusters developed in our study evolved over time according to EW. Cluster 3 was the most severe cluster, showing an increase in the number of patients during the second pandemic wave, which could be explained by the low availability of medical oxygen in the Peruvian territory during both the first and second waves of COVID-19 (Fraser, 2020; Herrera‐Añazco et al., 2021; Schwalb and Seas, 2021). Additionally, cluster 2 remained constant over time, whereas cluster 4 predominated in the first pandemic wave and then significantly declined.

Cluster 1 did not show a characteristic clinical pattern in comparison with the other groups evaluated. This cluster included the largest number of individuals and with the highest proportion of females and was the most predominant in the two pandemic waves evaluated in this study. With respect to age, this cluster was characterized by the highest proportion of children and adolescents between 5 and 17 years of age, as well as young people and adults. These results are similar to those described in a cluster study of COVID-19 patients by a French group in which their cluster 3 was made up by young patients with gastrointestinal symptoms and had the highest survival rate (96.5%) (Bondeelle et al., 2021).

Cluster 2 presented a high frequency of patients with risk factors for COVID-19 including cardiovascular disease, diabetes, obesity, and pulmonary disease, and was only surpassed by the "severe" cluster 3. These results partially coincide with a cluster study conducted in France, which showed a higher proportion of systemic, respiratory, and gastrointestinal symptoms in cluster 3, characterized by the highest survival rate in oncohematology patients (Bondeelle et al., 2021). Our findings also partially coincide with those of study performed in China, in which cough was predominant in group B, classified as "intermediate-severe." However, the presence of respiratory symptoms, gastrointestinal symptoms, and systemic symptoms was higher in the group A or "severe" group (Han et al., 2021). Our results coincide with those described in a similar study conducted in Spain, in which cluster C4 was characterized by diarrhea, vomiting, and abdominal pain, while cluster C3 was dominated by headache, sore throat, and arthromyalgia (Rubio-Rivas et al., 2020). Finally, cluster 2 of our study presented the highest frequency of anosmia and ageusia, similar to that described mainly in cluster C2 in a Spanish cluster analysis study (Rubio-Rivas et al., 2020).

Cluster 3 was characterized by dyspnea, the most severe sign among the four groups, and thus, this cluster was described as the "severe" group. This cluster is similar to what was described in an oncohematological hospital in France, in which a cluster analysis was performed in COVID-19 patients, finding that the majority of patients presented dyspnea, mainly in cluster 3 (88%) and this sign differed among clusters, making it important to address as a serious prognostic factor (Bondeelle et al., 2021). In a study performed in China, the second-highest proportion of dyspnea was found in patients in cluster A with greater severity (35.2%) (31). This finding is also consistent with that described in a cluster analysis of just over 12,000 patients hospitalized for COVID-19 in Spain, in which cluster C1 accounted for the highest in-hospital mortality, and was characterized by being the largest cluster and comprising the triad of fever, cough and dyspnea (34).

The cluster 3 was dominated by the largest number of individuals with comorbidities, similar to what was described by Bondeelle et al. who identified a higher proportion of comorbidities in patients grouped in cluster 2, characterized by the lowest survival rate (50.6%) (Bondeelle et al., 2021). Additionally, our findings coincide with those of a study carried out in the United States following a cluster analysis, which found more than 50% of patients with the lowest proportion of SARS-CoV-2 positivity without comorbidities, concluding that presenting comorbidities represents a risk factor for severe COVID-19. (35).

Cluster 4 included the highest number of pregnant women and children ranging in age from 0 to 4 years old, which could explain the fact that this group was also characterized by the predominance of being asymptomatic. In addition, this cluster had the lowest number of individuals with severe systemic and gastrointestinal disease. Nevertheless, this cluster had the second highest number of patients with renal disease and cancer, surpassed only by cluster 4. This is similar to what was found in a French study, in which only 4% of patients in cluster 1 had acute renal failure (Bondeelle et al., 2021; Cholankeril et al., 2020; Pan et al., 2020; Tian et al., 2020; Zhang et al., 2020).

Limitations and strengths

This research has some limitations. First, there may be potential information bias because the clinical-epidemiological variables were not obtained from the clinical histories of the patients selected for the present study. In addition, it was not possible to obtain the measurement of biochemical parameters such as glucose, fibrinogen, and D-dimer (Mollinedo-Gajate et al., 2021). Nor was it possible to analyze the variation in pharmacological therapy (use of anticoagulants, corticosteroids, etc.) administered to patients with COVID-19 infection (12) and, particularly in the cluster analysis, it was not possible to obtain measurements of severe outcomes such as hospitalization in critical units and death, biochemical markers, hospital stay, among others. Second, there may be a selection bias, given that our findings cannot be inferred to the entire study population of patients attended in the first and second waves of the pandemic in the Lambayeque region, or at the national level, since the data were from patients attended only in the EsSalud Lambayeque health care network. Third, due to the cross-sectional design of the study, it is not possible to attribute causality among the clinical and epidemiological characteristics that were associated with the first and second pandemic waves. Nevertheless, this study analyzed a large and diverse sample of patients treated for COVID-19 in the health care networks of the Lambayeque region and the findings obtained allow the development of future studies aimed at identifying the clinical-epidemiological variation of COVID-19 patients at local and regional levels, including diagnostic and therapeutic variables to explain the most relevant factors between each pandemic wave.

Conclusions

In our study, COVID-19 patients were grouped into four clusters with a characteristic clinical pattern using CLARA-based cluster analysis. Cluster 2 was characterized by gastrointestinal and respiratory involvement. Cluster 3 was denominated the severe cluster due to the presence of a higher frequency of dyspnea and older adult patients with comorbidities. In addition, we determined that dyspnea, respiratory and gastrointestinal symptoms, and comorbidities were higher in the second than in the first wave of the pandemic.

Funding

This study was supported by the Instituto de Evaluación de Tecnologías en Salud e Investigación (IETSI), EsSalud, Peru.

Declaration of competing interest

The authors have no conflicts of interests to declare.

References

Abd El-Raheem GOH, Mohamed DSI, Yousif MAA, Elamin HES. Characteristics and severity of COVID-19 among Sudanese patients during the waves of the pandemic. Sci Afr 2021;14:e01033. https://doi.org/10.1016/j.sciaf.2021.e01033.

Area I, Lorenzo H, Marcos PJ, Nieto JJ. One Year of the COVID-19 Pandemic in Galicia: A Global View of Age-Group Statistics during Three Waves. Int J Environ Res Public Health 2021;18. https://doi.org/10.3390/ijerph18105104.

Boletín Epidemiológico del Perú. Volumen 30-SE 18. Centro Nacional de Epidemiología, Prevención y Control de Enfermedades (CDC-Perú). n.d.

Bondeelle L, Chevret S, Cassonnet S, Harel S, Denis B, de Castro N, et al. Profiles and outcomes in patients with COVID-19 admitted to wards of a French oncohematological hospital: A clustering approach. PLoS One 2021;16:e0250569. https://doi.org/10.1371/journal.pone.0250569.

Cholankeril G, Podboy A, Aivaliotis VI, Tarlow B, Pham EA, Spencer SP, et al. High Prevalence of Concurrent Gastrointestinal Manifestations in Patients With Severe Acute Respiratory Syndrome Coronavirus 2: Early Experience From California. Gastroenterology 2020;159:775–7. https://doi.org/10.1053/j.gastro.2020.04.008.

Díaz-Vélez C, Failoc-Rojas VE, Valladares-Garrido MJ, Colchado J, Carrera-Acosta L, Becerra M, et al. SARS-CoV-2 seroprevalence study in Lambayeque, Peru. June–July 2020. PeerJ 2021a;9:e11210. https://doi.org/10.7717/peerj.11210.

Díaz-Vélez C, Fernández-Mogollón J, Neciosup-Puicán E, Colchado I, Ortiz-Millones J, Becerra-Torres M, et al. Experiencia de la estrategia implementada por los equipos de respuesta rápida y seguimiento clínico para reducir la letalidad por COVID-19, Lambayeque-Perú 2020. Revista del Cuerpo Médico Hospital Nacional Almanzor Aguinaga Asenjo 2021b;14:410–7. https://doi.org/10.35434/rcmhnaaa.2021.143.1285.

Díaz-Vélez C, Urrunaga-Pastor D, Romero-Cerdán A, Peña-Sánchez ER, Fernández Mogollon JL, Cossio Chafloque JD, et al. Risk factors for mortality in hospitalized patients with COVID-19 from three hospitals in Peru: a retrospective cohort study. F1000Res 2021c;10:224. https://doi.org/10.12688/f1000research.51474.1.

Fan G, Yang Z, Lin Q, Zhao S, Yang L, He D. Decreased Case Fatality Rate of COVID-19 in the Second Wave: A study in 53 countries or regions. Transbound Emerg Dis 2021;68:213–5. https://doi.org/10.1111/tbed.13819.

Fraser B. COVID-19 strains remote regions of Peru. Lancet 2020;395:1684. https://doi.org/10.1016/S0140-6736(20)31236-8.

Friston KJ, Parr T, Zeidman P, Razi A, Flandin G, Daunizeau J, et al. Second waves, social distancing, and the spread of COVID-19 across the USA. Wellcome Open Res 2021;5. https://doi.org/10.12688/wellcomeopenres.15986.3.

Griffin DO, Brennan-Rieder D, Ngo B, Kory P, Confalonieri M, Shapiro L, et al. The Importance of Understanding the Stages of COVID-19 in Treatment and Trials. AIDS Rev 2021;23:40–7. https://doi.org/10.24875/AIDSRev.200001261.

Guo Y-R, Cao Q-D, Hong Z-S, Tan Y-Y, Chen S-D, Jin H-J, et al. The origin, transmission and clinical therapies on coronavirus disease 2019 (COVID-19) outbreak - an update on the status. Mil Med Res 2020;7:11. https://doi.org/10.1186/s40779-020-00240-0.

Han L, Shen P, Yan J, Huang Yao, Ba X, Lin W, et al. Exploring the Clinical Characteristics of COVID-19 Clusters Identified Using Factor Analysis of Mixed Data-Based Cluster Analysis. Front Med (Lausanne) 2021;8:644724. https://doi.org/10.3389/fmed.2021.644724.

Herrera‐Añazco P, Uyen‐Cateriano A, Mezones‐Holguin E, Taype‐Rondan A, Mayta‐Tristan P, Malaga G, et al. Some lessons that Peru did not learn before the second wave of COVID‐19. Int J Health Plann Manage 2021:10.1002/hpm.3135. https://doi.org/10.1002/hpm.3135.

Huatuco-Hernández JA, Paredes-Villanueva FJ, Quispe-Cuestas MI, Fiestas-Pflücker GA, Nuñez-Rodas M, Salazar–Cuba X, et al. Características maternas y resultados perinatales en mujeres peruanas infectadas con COVID-19: Un estudio observacional y transversal. Revista del Cuerpo Médico Hospital Nacional Almanzor Aguinaga Asenjo 2021;14:344–51. https://doi.org/10.35434/rcmhnaaa.2021.143.1276.

Iftimie S, López-Azcona AF, Vallverdú I, Hernández-Flix S, Febrer G de, Parra S, et al. First and second waves of coronavirus disease-19: A comparative study in hospitalized patients in Reus, Spain. PLOS ONE 2021;16:e0248029. https://doi.org/10.1371/journal.pone.0248029.

Instituto de Evaluación de Tecnologías en Salud e Investigación. Guía de Práctica Clínica para el Manejo de COVID-19: Guía en Versión Extensa. Versión 1, mayo 2021. 2021.

Jiang F, Deng L, Zhang L, Cai Y, Cheung CW, Xia Z. Review of the Clinical Characteristics of Coronavirus Disease 2019 (COVID-19). J Gen Intern Med 2020;35:1545–9. https://doi.org/10.1007/s11606-020-05762-w.

Kassambara A. Practical Guide to Cluster Analysis in R: Unsupervised Machine Learning. STHDA; 2017.

Mocanu A, Noja GG, Istodor AV, Moise G, Leretter M, Rusu L-C, et al. Individual Characteristics as Prognostic Factors of the Evolution of Hospitalized COVID-19 Romanian Patients: A Comparative Observational Study between the First and Second Waves Based on Gaussian Graphical Models and Structural Equation Modeling. J Clin Med 2021;10. https://doi.org/10.3390/jcm10091958.

Mollinedo-Gajate I, Villar-Álvarez F, Zambrano-Chacón M de los Á, Núñez-García L, de la Dueña-Muñoz L, López-Chang C, et al. First and Second Waves of Coronavirus Disease 2019 in Madrid, Spain: Clinical Characteristics and Hematological Risk Factors Associated With Critical/Fatal Illness. Critical Care Explorations 2021;3:e0346. https://doi.org/10.1097/CCE.0000000000000346.

Mortality Analyses. Johns Hopkins Coronavirus Resource Center n.d. https://coronavirus.jhu.edu/data/mortality (accessed June 15, 2021).

Pan L, Mu M, Yang P, Sun Y, Wang R, Yan J, et al. Clinical Characteristics of COVID-19 Patients With Digestive Symptoms in Hubei, China: A Descriptive, Cross-Sectional, Multicenter Study. Am J Gastroenterol 2020;115:766–73. https://doi.org/10.14309/ajg.0000000000000620.

Peru: WHO Coronavirus Disease (COVID-19) Dashboard With Vaccination Data. n.d. https://covid19.who.int (accessed June 15, 2021).

Rubio-Rivas M, Corbella X, Mora-Luján JM, Loureiro-Amigo J, López Sampalo A, Yera Bergua C, et al. Predicting Clinical Outcome with Phenotypic Clusters in COVID-19 Pneumonia: An Analysis of 12,066 Hospitalized Patients from the Spanish Registry SEMI-COVID-19. J Clin Med 2020;9:E3488. https://doi.org/10.3390/jcm9113488.

Saito S, Asai Y, Matsunaga N, Hayakawa K, Terada M, Ohtsu H, et al. First and second COVID-19 waves in Japan: A comparison of disease severity and characteristics. J Infect 2021;82:84–123. https://doi.org/10.1016/j.jinf.2020.10.033.

Salyer SJ, Maeda J, Sembuche S, Kebede Y, Tshangela A, Moussif M, et al. The first and second waves of the COVID-19 pandemic in Africa: a cross-sectional study. Lancet 2021;397:1265–75. https://doi.org/10.1016/S0140-6736(21)00632-2.

Schwalb A, Seas C. The COVID-19 Pandemic in Peru: What Went Wrong? Am J Trop Med Hyg 2021;104:1176–8. https://doi.org/10.4269/ajtmh.20-1323.

Situación del COVID-19 en el Perú. CDC MINSA n.d. https://www.dge.gob.pe/portalnuevo/covid-19/covid-cajas/situacion-del-covid-19-en-el-peru/ (accessed June 15, 2021).

Soriano V, Ganado-Pinilla P, Sanchez-Santos M, Gómez-Gallego F, Barreiro P, de Mendoza C, et al. Main differences between the first and second waves of COVID-19 in Madrid, Spain. Int J Infect Dis 2021;105:374–6. https://doi.org/10.1016/j.ijid.2021.02.115.

Tian Y, Rong L, Nian W, He Y. Review article: gastrointestinal features in COVID-19 and the possibility of faecal transmission. Aliment Pharmacol Ther 2020;51:843–51. https://doi.org/10.1111/apt.15731.

Vinceti M, Filippini T, Rothman KJ, Di Federico S, Orsini N. SARS-CoV-2 infection incidence during the first and second COVID-19 waves in Italy. Environ Res 2021;197:111097. https://doi.org/10.1016/j.envres.2021.111097.

WHO Coronavirus (COVID-19) Dashboard. n.d. https://covid19.who.int (accessed June 15, 2021).

Zhang X, Tang C, Tian D, Hou X, Yang Y. Management of Digestive Disorders and Procedures Associated With COVID-19. Am J Gastroenterol 2020;115:1153–5. https://doi.org/10.14309/ajg.0000000000000728.

CRediT authorship contribution statement

Mario J. Valladares-Garrido: Data curation, Formal analysis, Investigation, Methodology, Writing – original draft, Writing – review & editing. Virgilio E. Failoc-Rojas: Data curation, Investigation, Methodology, Writing – original draft, Writing – review & editing. Percy Soto-Becerra: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Writing – original draft, Writing – review & editing. Sandra Zeña-Ñañez: Investigation, Methodology, Writing – original draft, Writing – review & editing. J. Smith Torres-Roman: Investigation, Methodology, Writing – original draft, Writing – review & editing. Jorge L. Fernández-Mogollón: Investigation, Methodology, Writing – original draft, Writing – review & editing. Irina G. Colchado-Palacios: Investigation, Methodology, Writing – original draft, Writing – review & editing. Carlos E. Apolaya-Segura: Investigation, Methodology, Writing – original draft, Writing – review & editing. Jhoni A. Dávila-Gonzales: Investigation, Methodology, Writing – original draft, Writing – review & editing. Laura R. Arce-Villalobos: Investigation, Methodology, Writing – original draft, Writing – review & editing. Roxana del Pilar Neciosup-Puican: Investigation, Methodology, Writing – original draft, Writing – review & editing. Alexander G. Calvay-Requejo: Investigation, Methodology, Writing – original draft, Writing – review & editing. Jorge L. Maguiña: Conceptualization, Investigation, Methodology, Writing – original draft, Writing – review & editing. Moisés Apolaya-Segura: Conceptualization, Investigation, Methodology, Writing – original draft, Writing – review & editing. Cristian Díaz-Vélez: Investigation, Methodology, Supervision, Writing – original draft, Writing – review & editing.