The highest number of confirmed cases for one day coincided with outbreaks of local transmission in two institutions where vulnerable people lived (a nursing home and a social security institution where 47 and 58 cases were identified, respectively). In this period, small- and moderate-sized outbreaks were relatively frequent, mainly within family groups and involving some sort of violation of established preventive measures, as revealed in epidemiological followup.[8]

The trend line calculated by five-day moving averages (Figure 1) smooths out the short-term fluctuations derived from the time elapsed between identification of suspected cases or contacts of primary cases and the sampling, transportation, RT-PCR testing and the cutoff timing for including results in the national registry and reporting system. The confirmed curve shows a rise of just over a month in duration, followed by a plateau lasting for approximately 15 days. The descent begins near day 45 and is interrupted only briefly by spikes in cases associated with the outbreaks already mentioned. The weekly case record corroborates our earlier assertion that the curve has a 5-week ascending arm and then begins to descend for the next 11 weeks (Figure 2).

During the 110-day period studied, the 2340 people infected with the SARS-CoV-2 virus constituted a national incidence rate of 20.7 per 100,000 population. Of this total, 1174 were men (50.2%) and 1166 women (49.8%); 162 (7.0%) acquired the infection abroad, mainly in the USA and Spain, countries with significant travel to and from Cuba. The largest number of cases sorted according to geographic area occurred in western Cuba, specifically in Havana, the capital and most cosmopolitan city. Havana alone accounted for 54.7% of the country’s total cases.

As a result of contact tracing, 88.4% of confirmed cases were traced to previously confirmed cases, and 54% were identified when asymptomatic or presymptomatic, suggesting that most were found in the early stages of infection. This has probably contributed to reducing SARS-CoV-2 transmission in Cuba, as contacts that are traced from previously confirmed cases are referred for quarantine in isolation centers for the maximum incubation period counted from the most recent exposure that contact had with their index case.[7]

Total confirmed cases included persons from 1 year to 101 years of age. For the purposes of this analysis, they were divided into the following age groups: <20 (12.4%); 20–39 (29.9%); 40–59 (35.6%); and ≥60 (22.1%) years.

In the 110-day period studied COVID-19 was the cause of 86 deaths, resulting in case fatality rate of 3.7%. This rate in Cuba was lower than that reported worldwide (4.96%) and lower than that of Latin America (4.8%) in the same time frame. The monthly case fatality rate decreased to 1.2% in June. Of the 86 deceased, 52 (60.5%) were men and 34 (39.5%) women, with case fatality considerably higher for men (4.4% vs. 2.9%), despite the fact that the average age of men was lower (70 years vs. 79.1 years). The deceased included two non-Cubans (one Russian and one Italian), and a Cuban national residing in Spain, all of whom were infected abroad.

Of the deceased, 85% had a history of one or more serious chronic non-communicable disease (NCD) comorbidities. Table 1 summarizes the most common NCDs associated with SARS-CoV-2 infection in the 86 people who died. High blood pressure and diabetes mellitus were the most frequent, which concurs with reports in the international literature.[9] Given the wide range of comorbidities that sometimes accompany positive cases it is not possible to assume that COVID-19 was the primary cause of death in all cases, although older adults and people of any age with serious underlying conditions are known to be at increased risk of becoming seriously ill and dying if infected with SARS-CoV-2.[10]

Figure 3 shows the results of estimating the effective reproductive number R_t using daily data from new cases reported between March 11 and June 28, 2020 across Cuba. The dark line corresponds to the mean of each value and the shaded lines correspond to 95% confidence intervals. R_t expresses the epidemic’s reproduction ratio; a key goal for stemming any epidemic is to achieve R_t <1, indicating that infected individuals are not causing new cases. Both R_t and R₀ by themselves are insufficient measures to characterize the dynamics of any disease in a population, but they have value as a complement to other indicators, as illustrated in the Cuban experience.

Researchers from the University of Havana’s Mathematics Department applied several prognostic models for the course of the epidemic in Cuba (Figure 4). The SIR model (Susceptible-Infected-Recovered), based on a system of ordinary differential equations, has been used elsewhere and was chosen due to its simplicity and ease of interpretation.[14,15] Figure 4 shows the expected case curves in three theoretically possible scenarios (favorable, moderate and critical). The active confirmed cases reported each day are registered in black (total cases minus those recovered or deceased).

The Cuban strategy included active screening for persons with asymptomatic infections. This was supported by the use of an epidemiological surveillance system (which tracks the location and contacts of potential vectors of communicable diseases) in place throughout the country for many years before the appearance of COVID-19.[7,16] This explains why 54% of confirmed cases were either in asymptomatic or presymptomatic stages at the time of their confirmation, and the share of asymptomatic cases continued to increase as the epidemic evolved, reaching close to 70% in the final weeks of the 110-day period.

Surveillance of international travelers was established before the first cases were confirmed in Cuba, and two weeks after these confirmations (on March 24, 2020), the decision was made to close airports to international flights.[17] Before closing the airports on March 24, contacts of COVID-19–positive cases (declared ‘suspected cases’) were advised to remain quarantined for 14 days in hospital centers outfitted for this purpose, or were monitored at home for the same period by primary healthcare personnel, restricting movement outside and maintaining physical distancing. This made it possible to trace all contacts of the first detected case. All these contacts underwent RT-PCR tests between the third and fourth days of their quarantine, and those with symptoms were referred to hospital centers with a higher level of clinical capacity.[7] At no time during the 110 days did COVID-19 patient care cause either primary care facilities or hospital emergency services to collapse, and intensive-care unit (ICU) capacities were always available. At the end of this period, a total of 39 infected persons were still in hospital with COVID-19, one of them in the ICU.

It is known that SARS-CoV-2 infection can spread rapidly in hospital settings.[18,19] This was also the case in Cuba, where 291 hospital workers were infected (12.1% of cases), including 212 doctors and nursing staff, with no deaths among them.

To achieve physical and social distancing at the population level, government authorities eliminated large public gatherings, closed schools and non-essential workplaces, and limited public transportation, among other measures, recommending in a massive public information campaign that people wash or disinfect hands regularly, use face masks outside and leave home only when necessary. Public messaging emphasized the importance of these habits for all individuals, especially for high-risk groups. Measures were more restrictive in communities with local transmission outbreaks or a large number of cases.[17] Various researchers have concluded that such strategies are effective for cutting the chain of person-to-person transmission of SARS-CoV-2. All of these interventions are included in WHO guidelines for confronting the pandemic[12] and in the Oxford COVID-19 Government Response Tracker (OXCGRT).[20] OXCGRT was developed by Oxford University (UK) as a tool to assess government response to the pandemic which achieved and maintained a rating of 100 since May 11, 2020 (ratings are assigned from 1 to 100, based on an aggregate of 17 indicators measuring government response to the pandemic).[20]

The Cuban strategy has also incorporated nationwide door-to-door screening for persons presenting with fever and/or respiratory symptoms, carried out by primary care professionals supported by medical sciences students.[7] Additionally, an application (‘app’) for mobile devices (a ‘virtual screener’) was developed for people to self-evaluate and indicate if they present symptoms, which then advises local health authorities so that a primary healthcare service provider can visit them at home.[21] Both components of the strategy have provided valuable complementary epidemiological information on presence of possible clinical cases of COVID-19 and have opened up new possibilities for surveillance and control of other diseases in the future.

Research is continuing to identify asymptomatic cases in various population groups in response to WHO’s call for population-based serological studies at the local or national level.[22] A national infection prevalence survey based on infection research (RT-PCR and antibody testing) is also being conducted in a probabilistic sample representative of the entire Cuban population that includes 4000 people of all ages and 1300 households.

The main epidemiological indicators associated with morbidity (number of confirmed cases, incidence rate, doubling time of cases, contacts traced, rate of tests per million population) have been consistent with the drop in infections made possible by efforts to detect both asymptomatic and symptomatic individuals and follow-up with their contacts. Thus, more than 85% of confirmed cases were contacts of previous cases. The level of virus penetration and the population’s immune level to SARS-CoV-2 is still unknown but may be elucidated by the serological study currently being conducted in a representative sample of the population.

The >14,000 RT-PCR tests carried out per million population (pmp) is considered a good indicator of case finding, especially since it has been guided by active case detection carried out at the primary care level for both asymptomatic contacts as well as persons with respiratory symptoms. However, the country continues to work to obtain resources to increase this indicator (pmp) to the extent that the complex economic situation allows.

Active nationwide door-to-door screening for symptomatic cases through use of the primary healthcare system and medical schools, admission of all positive cases to hospitals and testing of all their known contacts are typical characteristics of the Cuban response to the COVID-19 epidemic. All these ensured that necessary care arrived in a timely manner, before complications. These measures may also have contributed to the relatively low case fatality rate that was observed in the country, as most fatalities occurred in persons with an average age of 74 years, almost all of whom (85%) had concurrent underlying NCDs. This indicator (case fatality rate) was lower than the global average and the average for countries of the western hemisphere calculated within the same date range.[24] Additionally, the percentage of deaths among seriously ill patients and among critically ill patients (those requiring respiration assistance), at less than 20%, is much lower than that reported elsewhere in the literature, which ranges between 50% and 70%.[7,10]

Available epidemiological indicators suggest that timely introduction of social distancing (begun on March 24, 2020 in Cuba) is recognized as an effective means of reducing transmission. This may help explain Cuba’s favorable results during this period in flattening the curve of infection.

The timing of measures adopted and the intensity with which they were applied decreased opportunities for transmission between contacts, which in turn contributed to reaching the peak of the epidemic faster and with a relatively lower number of cumulative deaths during this period.

In the database available from Johns Hopkins University, USA, comparing rates of infection near epidemic day 60 in 10 Latin American countries revealed that the only country exhibiting an already flattened curve was Cuba. Costa Rica and Uruguay had begun to stabilize their curves at this point, while Peru and Brazil had the highest rates of multiplying cases.[25]

By the end of June 2020, 14 of the 15 provinces of Cuba and the special municipality of the Isle of Youth had reported no new cases for >15 days (in some provinces, for >30 days). This was one of several main criteria that allowed these territories to transition to the first of three recovery phases contemplated in the national strategy for the country’s gradual reopening. Havana was the last province to achieve that certification, beginning Phase 1 on July 3, 2020, the same day that the rest of the provinces (except Matanzas) entered Phase 2.[7]

Cuba’s response to the epidemic has been multisectoral.[25] The Ministry of Public Health, in charge of organizing Cuba’s free and universal health services at all levels of care, has assumed technical leadership. However, the role of Civil Defense, already a key player during other disasters and epidemics in Cuba, merits recognition. Another key component of the Cuban strategy has been the contribution by scientists and their institutions as advisors to government for the design and application of various strategies, ranging from the epidemiological to patient care and development of vaccine candidates.[13] Last but not least, an essential pillar of the Cuban strategy has been collaboration by the Cuban public in adopting the measures designed and implemented by the healthcare system and government authorities.