High coverage COVID-19 mRNA vaccination rapidly controls SARS-CoV-2 transmission in Long-Term Care Facilities

Abstract

Residents of Long-Term Care Facilities (LTCFs) represent a major share of COVID-19 deaths worldwide. Measuring the vaccine effectiveness among the most vulnerable in these settings is essential to monitor and improve mitigation strategies.

We evaluated the early effect of the administration of BNT162b2 mRNA vaccines to individuals older than 64 years residing in LTCFs in Catalonia, a region of Spain. We monitored all the SARS-CoV-2 documented infections and deaths among LTCFs residents from February 6th to March 28th, 2021, the subsequent time period after which 70% of them were fully vaccinated. We developed a modeling framework based on the relation between community and LTFCs transmission during the pre-vaccination period (July -December 2020) and compared the true observations with the counterfactual model predictions. As a measure of vaccine effectiveness, we computed the total reduction in SARS-CoV-2 documented infections and deaths among residents of LTCFs over time, as well as the reduction on the detected transmission for all the LTCFs.

We estimated that once more than 70% of the LTCFs population were fully vaccinated, 74% (58%−81%, 90% CI) of COVID-19 deaths and 75% (36%−86%, 90% CI) of all expected documented infections among LTCFs residents were prevented. Further, detectable transmission among LTCFs residents was reduced up to 90% (76–93%, 90%CI) relative to that expected given transmission in the community.

Our findings provide evidence that high-coverage vaccination is the most effective intervention to prevent SARS-CoV-2 transmission and death among LTCFs residents. Conditional on key factors such as vaccine roll out, escape and coverage --across age groups--, widespread vaccination could be a feasible avenue to control the COVID-19 pandemic.

Introduction

Widespread vaccination has the potential to significantly reduce SARS_CoV-2 infections and deaths, and subsequently improve social and economic conditions [1]. Available mRNA COVID-19 vaccines have been approved due to their capacity to reduce symptomatic disease, hospitalizations, and deaths in clinical trials [2,3]; evidence of their real-world effectiveness is growing [4–6], but confirmation still remains limited to certain populations and settings [7–9].

Among all populations, residents of Long-Term Care Facilities (LTCFs) represent a major share of COVID-19 deaths, with a 7-fold higher incidence of death compared to the general population in the US [10,11]. As a consequence, they have been prioritized for vaccinations in most settings. While observational data post-vaccination are still under evaluation at the time of this work, particularly regarding residents of LTCFs [4,12], early assessments on whether clinical trial results are good indicators of vaccine effectiveness in LTCFs would help refine control strategies [1].

In this study, we aimed to quantify the early effect of the administration of the BNT162b2 mRNA COVID-19 vaccine on reducing the risk of SARS-CoV-2 transmission and COVID-19 death among residents of LTCFs in Catalonia (Spain), where high rates of full vaccination among individuals older than 64y (>90% coverage) were reached around 3 months after vaccinations began in December 27, 2020. Prior to the vaccination campaign, the control of SARS-CoV-2 transmission in LTCFs in Catalonia [13] relied on: a) protocolized prevention measures at the individual and facility level, and b) rigorous and timely case ascertainment (including passive case and active case detection) and isolation standards. Protocols regulating the conditions of external visits to the facilities, as well as screening protocols for exits/entries of residents were tightened between December-January [14] but were relaxed once the vaccination campaign was completed. Further, restrictions on individuals’ mobility were implemented by the Catalonian Government at different levels through the territory based on epidemiological risk; in all of Spain, the tightest restrictions were applied in March-June 2020 (which included a country-wide shelter-in-place intervention leading to control of SARS-CoV-2 transmission) and in December 2020-January 2021 [15].

Methods

The target population analysed in this work was all individuals older than 64y living in care homes in Catalonia, estimated to be around 58,000 in total (see details in supplementary information section S1), between July 2020 and March 2021.This population was vaccinated using the BNT162b2 mRNA COVID-19 vaccine following the guidelines of the Spanish Ministry of Health. The epidemiological data used in this study was collected from --a publicly available repository provided by-- the Health Department depending on the Generalitat de Catalunya, the Government of Catalonia. Data on LTCF are collected and updated on a daily basis using health reports from the Primary Care Clinical Station (Care home census), Aggregated Care Health Register (PCR results and deaths), Orfeu, the program for registering virological test results in care homes, and the Catalan Shared Clinical Record where vaccination are registered.

We defined three COVID-19 outcomes to evaluate vaccine efficacy: a) documented infections, comprised of all new infections reported during the study period, independent of symptoms and vaccination status, b) documented deaths, comprised of all deaths attributable to COVID-19 reported during the study period, also independent of vaccination status, and c) detected county-level transmission (herein detected transmission, used as a binary indicator of transmission) defined as at least one documented infection in any facility within a county per unit of time. Each case was assigned the date of diagnostic or testing as the reporting date. LTCFs residents were defined as those living in a LTCF and older than 64y. The general population, or “community”, was defined as all people in a specific area not living in LTCFs. All vaccinated individuals received the two doses of the BNT162b2 mRNA vaccine. Documented COVID-19 infections in LTCFs over time, identified with a molecular test (PCR or antigen test), were assumed to capture most infections given the surveillance protocols in place. Beginning in July 2020, the ascertainment of cases in LTCFs in Catalonia included symptomatic surveillance, tight outbreak investigation and testing of all contacts upon identification of one single case, as well as regular screening of all individuals and staff of a facility independently of symptoms (see further details on supplementary information, Section S1). Documented deaths attributable to COVID-19 included those with laboratory confirmation and those only meeting clinical and epidemiologic criteria. We used 3 spatial resolutions for our analysis, determined by the level of aggregation in the data: a) county level, which corresponds to the definition and boundaries of each “comarca” (n= 41) b) health care area level, which corresponds with the definition and boundaries of each “regió sanitaria” (n=9), and c) regional level, which refers to the largest spatial resolution corresponding to the whole Autonomous Community of Catalonia. For further details see Supplementary Information Section S1.

We generated multiple time series of daily confirmed infections, deaths, and vaccinations in LTCFs, aggregated by healthcare area level, and by the regional (highest aggregation) level. Similarly, we collected daily confirmed infections in the general population, at the healthcare-area level and regional level. For each of these time series, we took a moving weekly average around each day to smooth the daily variation of reporting. Further, we generated a time series of the detected county-level transmission by week. The pre-vaccination period was from July 6 to December 27 2020, when vaccination in LTCFs began. We evaluated the impact of vaccines during an evaluation time period from February 6 to March 28, 2021, the subsequent time period after which 70% of residents were vaccinated with two doses. The 70% threshold was chosen to represent the estimated herd immunity - the estimated level of immunity in a population that prevents uncontrolled spread of infections [16]. As a sensitivity test, we evaluated the effect of partial vaccination in a longer period starting on January 14, when 70% of residents had received the first dose of the vaccine.

We built regression models to predict the number of infections and deaths among LTCFs residents using community infections as inputs. These models were calibrated during the pre-vaccination period and then used to generate predictions in the absence of vaccines during the evaluation period. We compared the models’ counterfactual predictions with observations; discrepancies were used to quantify the effects of the vaccine. Further, for each county with at least one pre-vaccination week with a transmission event in LTCFs and at least one week without one (n=36), we built a logistic regression model to estimate the probability that at least one documented transmission would be observed among LTCFs residents in a given week, using community infections as input. These models were trained on the pre-vaccination period and then used to predict transmission events in the evaluation period among LTCFs residents; the aggregated county-level predicted probabilities and the aggregated observed transmissions were compared to quantify vaccine effectiveness in LTCFs. See Supplementary Information Section S1 for further details on the models.

Results

We quantified the effect of administering the BNT162b2 vaccine (a) on reducing deaths and documented infections among LTCFs residents older than 64y, and (b) on reducing detected transmission caused by SARS-CoV-2 in LTCFs in Catalonia.

Figure 1 A shows the temporal evolution of infections documented in the community and in LTCFs between July 6, 2020 – March 28 2021. For context, the cumulative vaccination coverage among all LTCFs residents is shown in panel B. Vaccination was deployed among residents and healthcare workers at similar times across LTCFs facilities in the region, beginning December 27 and reaching more than 95% of 2-dose coverage within 2 months.

Figure 1.

A) Comparison of the total community (grey) and LTCFs documented infections (red) trajectories in Catalonia, Spain. B) First and second dose vaccine coverage among LTCFs residents

Figure 2 A and B show predictions and observations of documented infections and deaths in all of Catalonia over time. We estimated that between February 6-March 28 2021, vaccines prevented 75% of documented infections (36% – 86%, 90% CI), and 74% of deaths (58% – 81%, 90% CI). As well, our analysis shows that two weeks after 70% of residents were fully vaccinated, detected transmission was significantly reduced by 69% (24–80% 90%CI), 54% (0–70%), 50% (0–68%), 69% (25–80%), and 90% (76–93% 90%CI) for each subsequent epidemiological week (Figure 2C).

Figure 2:

The predictions for infections (A) and deaths (B) across all of Catalonia. The solid lines show the model predictions from training July 6, 2020 through December 27, 2020, the darker shaded background shows the 50% prediction intervals (PI) and the lighter background shows the 90% PI. Vertical lines show key analysis time points: when vaccination started (solid), when 70% of residents received the first dose and when 70% of residents received the second dose. (C) The ratio between observed and predicted transmission at county level in Catalonia, represented by point estimates, grey for the training period and green for the prediction period; grey horizontal ribbon represents the 90% confidence interval. Solid green areas represent the prediction periods after vaccination starts.

Discussion

In this study we showed that high vaccination coverage (over 70%) prevented around 3 out of 4 expected COVID-19 deaths among residents of LTCFs in subsequent weeks, which is consistent with the vaccine effect on disease severity observed in clinical trials [2] and mortality reduction in other observational studies [6,12]. Further, we found a reduction in transmission after high vaccination coverage was reached while there still was transmission risk (spilled-over from the mainly unvaccinated community transmission), which is caused by both a reduction in vaccinated individuals’ probability of getting infected and a reduction in their probability of transmitting the virus [17].

LTCFs represent enclosed populations, where transmission is caused both by external introduction of the virus, mostly by the staff [18], and internal transmission. Observational findings from facilities with high infection-ascertainment, such as those studied here and elsewhere [4], may provide valuable information of what may be expected to happen in more general settings and populations. Given this, we conclude that beyond the specific risk factors of the population in LTCFs, such as age or comorbidities, high-coverage vaccination can rapidly control SARS-CoV-2 transmission in an enclosed population. Particularly, our analysis suggests that transmission is reduced 3–10-fold one month after vaccination has reached 70% coverage. Of note, our estimates of infections and deaths do not differentiate between infections in vaccinated or unvaccinated individuals, and therefore can be interpreted as the population-level effect of vaccination.

Our methodological assumptions may lead to underestimation of the true vaccine effect. First, our definition of transmission events at the county-level does not capture the facility-level transmission; thus, our results may fail to capture more dramatic facility-level reductions in transmission. Second, the counterfactual estimates of deaths produced by our models may be underestimations of mortality during the evaluation period. Indeed, the documented deaths were generally higher than those estimated by our models in December and January (Figure 2B). It is plausible that factors not considered in the model, such as seasonality [19] and/or the spread of more lethal variants [20] may have increased COVID-19 mortality in recent times. Further, as per guidelines, vaccinations in facilities with ongoing transmission were delayed, which would again lead to underestimation of the vaccine effectiveness. On the other hand, it is possible that relaxing of case ascertainment after the vaccination campaign could lead to overestimation of the effect on documented infections and detected transmission (but unlikely for documented deaths); therefore, we restricted our analysis to the period just after vaccination.

We acknowledge that our investigation has limitations. We assume that the rigorous screening standards in LTCFs in Catalonia led to infection ascertainment close to 100% and also that the ascertainment of community infections does not change dramatically over time. While this may appear unrealistic, public health authorities in Catalonia substantially increased the infection screening efforts in LTCFs before the vaccination campaign [21], beginning in December 2020.

Our model is based on the assumption that the time between disease transmission and case identification is not significantly different between individuals in the community and those in LTCFs. Further, regression models were not designed for accurate infection (or deaths) forecasting and as such, they may not fully capture the epidemiological dynamics, such as changes on COVID-19 restrictions policy over time. However, our efforts were focused on proper inference of the expected epidemic trajectory in the absence of vaccination. This goal is achieved as our models appear to reasonably capture the overall dynamics during the pre-vaccination time periods, even at high spatial granularity (Supplementary Information Figure S1–S2).

Finally, there could be unmeasured confounders, such as behavior or policy changes, not captured by our models, that may have changed the dynamics of transmission between the community and LTCFs over time --this may be the case in the health area Alt Pirineu i Aran prior to the beginning of the vaccination campaign (Supplementary Information Figure S1–S2). Nevertheless, tight restrictions on individuals’ mobility for the whole territory were homogeneously implemented beginning December 2020 and during the period of analysis in this work.

In spite of these limitations, our analyses provide evidence that vaccination may be the most effective intervention in controlling SARS-CoV-2 spread and subsequent risk of death available to date. If our findings continue to be confirmed by future studies, then, conditional on important factors such as vaccine roll out, escape and coverage [22,23], widespread vaccination could be shown to be a feasible avenue to control the COVID-19 pandemic.