Effect of COVID-19 vaccination on the SARS-CoV-2 transmission among social and household close contacts: A cohort study

Abstract

Background

COVID-19 vaccination was expected to reduce SARS-CoV-2 transmission, but the relevance of this effect remains unclear. We aimed to estimate the effectiveness of COVID-19 vaccination of the index cases and their close contacts in reducing the probability of SARS-CoV-2 transmission.

Methods

Transmission of SARS-CoV-2 infection was evaluated in two cohorts of adult close contacts of COVID-19 confirmed cases (social and household settings) by COVID-19 vaccination status of the index case and the close contact, from April to November 2021 in Navarre, Spain. The effects of vaccination of the index case and the close contact were estimated as (1–adjusted relative risk) × 100%.

Results

Among 19,631 social contacts, 3257 (17%) were confirmed with SARS-CoV-2. COVID-19 vaccination of the index case reduced infectiousness by 44% (95% CI, 27–57%), vaccination of the close contact reduced susceptibility by 69% (95% CI, 65–73%), and vaccination of both reduced transmissibility by 74% (95% CI, 70–78%) in social settings, suggesting some synergy of effects. Among 20,708 household contacts, 6269 (30%) were infected, and vaccine effectiveness estimates were 13% (95% CI, −5% to 28%), 61% (95% CI, 58–64%), and 52% (95% CI, 47–56%), respectively. These estimates were lower in older people and had not relevant differences between the Alpha (April-June) and Delta (July-November) variant periods.

Conclusions

COVID-19 vaccination reduces infectiousness and susceptibility; however, these effects are insufficient for complete control of SARS-CoV-2 transmission, especially in older people and household setting. Relaxation of preventive behaviors after vaccination may counteract part of the vaccine effect on transmission.

Introduction

SARS-CoV-2 can be transmitted from an infected person to their close contacts by droplets and air [1], [2], [3]. Since SARS-CoV-2 emerged in December 2019, control measures such as the use of facemasks, physical distancing, contact tracing, and isolation of cases helped to slow the spread [4], [5]. However, maintaining compliance with these measures over time is difficult, and their effects disappear as soon as the measures are suspended.

COVID-19 vaccination has been suggested as a preventive measure with long lasting effects [5]. Several studies found very high effectiveness of COVID-19 vaccination in preventing hospital admission and COVID-19 deaths, but effectiveness in preventing mild and asymptomatic SARS-CoV-2 infection was notably lower and declined progressively [6], [7]. Therefore, vaccinated individuals may acquire and transmit the SARS-CoV-2 infection [8], [9]. The effectiveness of COVID-19 vaccination in preventing SARS-CoV-2 transmission is crucial information for planning a prudent de-escalation of preventive measures in the environment of highly vulnerable people, following the increase in vaccination coverage [5], [10].

COVID-19 vaccination is expected to reduce SARS-CoV-2 transmission in the population by several mechanisms [11]. COVID-19 vaccines have demonstrated a considerable direct effect in preventing infection among vaccinated people (susceptibility) [7]. Vaccination may also reduce the infectiousness of breakthrough infections [10]. This reduced infectiousness may be related to decreased severity, symptom intensity, infectivity duration, and viral load [12]. Additionally, the indirect effect of the COVID-19 vaccines may reduce SARS-CoV-2 circulation in populations with high vaccination coverage [13].

SARS-CoV-2 transmission may happen in household and other settings, such as social meetings, workplaces, and schools [14], [15]. Close household contacts have been found to have a higher risk of infection than close contacts in other settings, since exposure is usually more intense and repeated in the household [16], [17], [18], [19], [20].

This study aimed to estimate the effectiveness of COVID-19 vaccination in reducing the infectiousness of index cases, the susceptibility of their close contacts, and the probability of SARS-CoV-2 transmission when both of them are vaccinated.

Material and methods

Study population and design

This study analyzed two prospective dynamic cohorts of adults (≥18 years old) based on the contact tracing of laboratory-confirmed COVID-19 cases in Navarre, Spain, where the Health Service provides universal health care, free at the point of service. A first cohort included social contacts and a second cohort included household contacts. These cohorts incorporated all adults covered by the Navarre Health Service, who had been close contacts of adult index cases with confirmed SARS-CoV-2 infection from April to November 2021. The study protocol was approved by the Ethical Committee for Clinical Research of Navarre (PI2020/45), which waived the requirement of obtaining signed consent since the study analyzed an anonymous database.

As a measure of infection control, all laboratory-confirmed COVID-19 cases were interviewed to identify their close contacts [21]. The index case was the first person who presented SARS-CoV-2 infection in each specific setting and was confirmed by reverse transcription-quantitative polymerase chain reaction (RT-qPCR) or antigen test. Only one index case per cluster was considered. Close contact of a COVID-19 index case was defined as any person who had face-to-face contact with a confirmed COVID-19 infected individual within 2 m for more than a total of 15 min without personal protection within a timeframe ranging from 2 days before to 10 days after the onset of symptoms of the case, or in the 2 days before to 10 days after the sample which led to confirmation was taken from asymptomatic cases [20], [22]. The same protocol was applied regardless of the contact setting. Close contacts were initially tested by RT-qPCR for SARS-CoV-2 in nasopharynx samples and 10 days after the last contact. A positive antigen test within 5 days from the symptom onset was also considered confirmatory of SARS-CoV-2 infection in symptomatic individuals, because this test had demonstrated very high specificity in these cases. Close contacts with negative antigen tests were retested with RT-qPCR. Contact tracing was registered including information on the index case, the close contact, and the risk exposure; and this information was electronically connected with the databases of laboratory-test results, medical records, and enhanced epidemiological surveillance of COVID-19. Close contacts non-covered by the Navarre Health Service, with a previous positive test for SARS-CoV-2, younger than 18 years, nursing home residents, and those who did not complete the testing protocol or who had been confirmed less than 3 days after the index case diagnosis date were excluded from the present study. This last exclusion pretended to reduce doubts about the directionality of transmission. Each close contact was only included once in the study. During the study period, there was high availability of diagnostic tests and personnel for testing and contact tracing of all confirmed cases.

The close contact variables were age group (18–34, 35–49, 50–59, 60–69, and ≥70 years), sex, major chronic conditions, the month of contact, and contact setting (social or household). Household close contacts were considered those who lived in the same house during the infectivity period of the index case, while the category of social setting included close contacts generated in any activity occurring outside the home, such as social life, workplace, educational settings, and others. The presence of symptoms in the index case was also considered.

The index cases samples with a cycle threshold ≤ 30 were tested by TaqPath™ COVID-19 RT-qPCR kit and TaqMan™ SARS-CoV-2 Mutation Panel (Thermo Fisher Scientific, USA) following the manufacturer’s instructions to get an approximation of the variant. S-gene target failure by TaqPath was used as a proxy for identifying the Alpha variant [23]. Among samples that tested positive for the S target, a second RT-qPCR (TaqMan) assay was used to detect possible cases of Delta variant (detection of the L452R mutation).

The COVID-19 vaccination campaign started on 27 December 2020 in Spain. Four vaccine products have been used: BNT162b2 mRNA (BioNTech-Pfizer, Mainz, Germany/New York, USA), mRNA-1273 (Moderna, Cambridge, US), ChAdOx1 nCoV-19 (Oxford-AstraZeneca, Cambridge, United Kingdom), and Ad26. COV2-S (Janssen-Cilag International NV, Beerse, Belgium). Vaccination was successively targeted from higher to lower age groups [24]. COVID-19 vaccine doses and date of administration of index cases and their close contacts were obtained from the regional vaccination register, and only registered doses were considered. Each dose was considered potentially effective 14 days after administration. Fully vaccinated people were considered 14 days after the first dose of Ad26. COV2-S or after the second dose of other vaccines.

Statistical analysis

Basic characteristics of both cohorts of close contacts were described. The secondary attack rate of SARS-CoV-2 infection in close contacts was compared by characteristics of the study population of close contacts and by COVID-19 vaccination status of the index case in both contact settings (social and household).

We used Cox regression models to evaluate the vaccine effectiveness in reducing infectiousness, susceptibility and transmissibility between index cases and their close contacts in social and household settings. The same risk period was assigned to everyone in the cohort; therefore, we estimated crude and adjusted relative risks (aRR) with 95% confidence intervals (CI). Adjusted models included age group, sex, major chronic condition of close contacts, vaccination status of the index case and the close contact, and month of contact.

To estimate the vaccine effectiveness, we used the combination of the vaccination status (unvaccinated, partially vaccinated, and fully vaccinated) of the index case and the close contact in nine categories. Unvaccinated close contact of an unvaccinated index case was the reference category. Although models included nine categories, to simplify the presentation of results, we only show the estimates of categories with partial vaccination in supplementary material, as this was a transitory status that affected a few subjects. The estimate of the category of fully vaccinated close contacts of unvaccinated index cases was considered as a measure of susceptibility, that of unvaccinated close contacts of fully vaccinated index cases was considered as a measure of infectiousness, and that of fully vaccinated close contacts of fully vaccinated index cases was considered as an indicator of transmissibility. Two periods were considered according to the dominant SARS-CoV-2 variant in the region: from April to June 2021 more than 90% of characterized cases were due to the Alpha variant, and from July to November 2021, more than 90% were due to the Delta variant.

The synergy between the vaccination status of the index case and their close contacts were evaluated by comparing the category of vaccinated contact of an unvaccinated index case with the category of vaccinated close contact of a vaccinated index case. Stratified analyses were performed by age groups (18–59 and ≥60 years), the presence of symptoms of the index case, the variant dominance period, and the variant detected in the index case (Alpha, Delta, and others).

The vaccine effectiveness was estimated as a percentage: (1 − aRR) × 100. P-values lower than 0.05 were considered statistically significant.

Results

Characteristics of the close contact cohorts

Among a total of 62,355 close contacts of 21,879 index case, 40,339 close contacts met the study inclusion criteria, 20,708 (51%) were close household contacts and 19,631 were close social contacts. As compared to the household contacts, social contacts less frequently were older than 50 years (38% versus 49%) and had received any COVID-19 vaccine (58% vs 67%); however, the close social contacts were more frequently contact of a vaccinated index case (36% vs 23%)( Fig. 1 and Table 1).

Fig. 1.

Flowchart of the cohort study of close contacts for evaluating vaccination effect on SARS-CoV-2 transmission in Navarre, Spain, April to November 2021. Contact means close contact; case means index case. n, number of close contacts under the specific condition.

Table 1.

SARS-CoV-2 infection in social and household close contacts by characteristics of the close contacts and vaccination status of the index case and the close contacts.

	Social contacts				Household contacts
Variables of the study population	Nº	%	Positive	SAR, %	Nº	%	Positive	SAR, %
Total	19,631	100	3257	17	20,708	100	6269	30
Age group, years
18–34	7312	37	1757	24	4081	20	1548	38
35–49	4894	25	619	13	6559	32	2184	33
50–59	2174	11	288	13	2297	11	708	31
60–69	2653	14	292	11	6206	30	1294	21
≥ 70	2598	13	301	12	1565	8	535	34
Sex
Male	9508	48	1692	18	9754	47	2922	30
Female	10,123	52	1565	16	10,954	53	3347	31
Major chronic condition
No	13,768	70	2350	17	14,963	72	4529	30
Yes	5863	30	907	16	5745	28	1740	30
Diabetes	960	5	140	15	961	5	331	34
Cancer	1257	6	157	13	1272	6	354	28
Liver cirrhosis	361	2	46	13	389	2	107	28
Renal disease	508	3	72	14	415	2	130	31
Immunodeficiency	146	1	17	12	143	1	55	39
Asthma	1535	8	273	18	1306	6	423	32
COPD	891	5	124	14	930	5	272	29
Cardiovascular disease	1546	8	235	15	1374	7	414	30
Dementia	85	0.4	13	15	72	0.3	16	22
Stroke	238	1	33	14	200	1	57	29
Rheumatic disease	162	1	23	14	198	1	56	28
Severe obesity	255	1	39	15	297	1	112	38
Vaccination status of close contact
Unvaccinated	8244	42	2058	25	6933	33	3138	45
Partially vaccinated	1873	10	201	11	2397	12	541	23
Fully vaccinated	9514	48	998	11	11,378	55	2590	23
Month of contact
April	3712	19	629	17	3469	17	1517	44
May	1513	8	182	12	1493	7	547	37
June	1146	6	162	14	957	5	300	31
July	5639	29	1347	24	7636	37	1523	20
August	2105	11	271	13	2805	14	771	28
September	660	3	66	10	565	3	196	35
October	1036	5	114	11	570	3	242	43
November	3820	19	486	13	3213	16	1173	37
Close contact of symptomatic index case
No	2324	12	259	11	3457	17	681	20
Yes	17,307	88	2998	17	17,251	83	5588	32
Close contact by vaccination status of the index case
Unvaccinated	12,631	64	2511	20	15,955	77	4777	30
Partially vaccinated	979	5	102	10	853	4	235	28
Fully vaccinated	6021	31	644	11	3900	19	1257	32

COPD, chronic obstructive pulmonary disease; SAR, secondary attack rate.

Among close social contacts, 3257 (17%) were confirmed with SARS-CoV-2, while among close household contacts the infection was detected in 6269 (30%). In both settings, the transmission was more frequent in close contacts younger than 35 years and unvaccinated, as well as when the index case was unvaccinated and presented symptoms (Table 1).

Vaccine effectiveness on SARS-CoV-2 transmission in social settings

The secondary attack rate from unvaccinated index cases to their unvaccinated close social contacts was 26% on average. Vaccination of the index case reduced infectiousness by 44% (95% CI 27–57%), vaccination of the close contact reduced susceptibility by 69% (95% CI 65–73%), and vaccination of both reduced transmissibility by 74% (95% CI 70–78%), suggesting an additional effect of vaccination of both compared to close contact vaccination alone (p = 0.015). These vaccine effectiveness estimates were similar in young adult contacts (42%, 95% CI 24–56%; 68%, 95% CI 62–73%; and 75%, 95% CI 70–79%; respectively), but suggested a lower reduction of susceptibility in contacts older than 60 years (35%, 95% CI −69% to 75%; 54%, 95% CI 30–70%; and 56%, 95% CI 32–71%; respectively) ( Fig. 2 and Supplementary Table S1).

Fig. 2.

Effect of COVID-19 vaccination on the SARS-CoV-2 transmission among social and household close contacts, overall and by age of the close contact. Contact means close contact; case means index case. ^a Vaccine effectiveness adjusted by age groups (18–34, 35–49, 50–69 and ≥70 years), sex, major chronic condition of close contacts, and month of contact.

The presence of symptoms in the index case was associated with higher secondary attack rates, regardless of the vaccination status of the index case and the close contact; however, vaccination of the close contacts appeared to be more protective in contacts of asymptomatic index cases than in contacts of symptomatic index cases (80% vs 68%). For all vaccination statuses, the secondary attack rates in close social contacts increased from the Alpha variant period (April-June) to the Delta variant period (July-November); however, for the same vaccination status, the vaccine effectiveness estimates were similar or slightly higher during the Delta variant period, with 47% reduction of infectiousness, 72% of susceptibility and 76% of transmissibility. Results of the analyses by dominant variant periods were consistent with those considering the variants detected in the index cases ( Table 2 and Supplementary Tables S2 and S3).

Table 2.

Effectiveness of COVID-19 vaccination on the SARS-CoV-2 transmission among close social contacts.

	Total nº	Positive nº	%	Crude RR (95% CI)	Adjusted RR (95% CI)a	Adjusted VE, % (95% CI)a
Contacts of symptomatic index case
Unvaccinated contact and case	6715	1793	27	1	1	Reference
Unvaccinated contact and vaccinated case	256	52	20	0.76 (0.58–1.00)	0.57 (0.43–0.76)	43 (24 to 57)
Vaccinated contact and unvaccinated case	3119	367	12	0.44 (0.39–0.49)	0.32 (0.28–0.38)	68 (62 to 72)
Vaccinated contact and case	5011	535	11	0.40 (0.36–0.44)	0.26 (0.23–0.31)	74 (69 to 77)
Contacts of asymptomatic index case
Unvaccinated contact and case	949	163	17	1	1	Reference
Unvaccinated contact and vaccinated case	57	8	14	0.82 (0.40–1.66)	0.56 (0.27–1.15)	44 (−15 to 73)
Vaccinated contact and unvaccinated case	536	33	6	0.36 (0.25–0.52)	0.20 (0.12–0.34)	80 (66 to 88)
Vaccinated contact and case	423	29	7	0.40 (0.27–0.59)	0.18 (0.10–0.32)	82 (68 to 90)
Contacts with major chronic condition
Unvaccinated contact and case	1745	441	25	1	1	Reference
Unvaccinated contact and vaccinated case	61	10	16	0.65 (0.35–1.21)	0.46 (0.24–0.88)	54 (12 to 76)
Vaccinated contact and unvaccinated case	1376	142	10	0.41 (0.34–0.49)	0.28 (0.21–0.37)	72 (63 to 79)
Vaccinated contact and case	1876	223	12	0.47 (0.40–0.55)	0.29 (0.22–0.38)	71 (62 to 78)
Contacts without major chronic condition
Unvaccinated contact and case	5919	1515	26	1	1	Reference
Unvaccinated contact and vaccinated case	252	50	20	0.78 (0.59–1.03)	0.58 (0.43–0.78)	42 (22 to 57)
Vaccinated contact and unvaccinated case	2279	258	11	0.44 (0.39–0.51)	0.33 (0.28–0.39)	67 (61 to 72)
Vaccinated contact and case	3558	341	10	0.37 (0.33–0.42)	0.24 (0.20–0.30)	76 (70 to 80)
April to June 2021
Unvaccinated contact and case	4811	834	17	1	1	Reference
Unvaccinated contact and vaccinated case	56	5	9	0.52 (0.21–1.24)	0.56 (0.23–1.36)	44 (–36 to 77)
Vaccinated contact and unvaccinated case	457	32	7	0.40 (0.28–0.58)	0.41 (0.29–0.60)	59 (40 to 71)
Vaccinated contact and case	30	0	0	NA	NA	NA
July to November 2021
Unvaccinated contact and case	2853	1122	39	1	1	Reference
Unvaccinated contact and vaccinated case	257	55	21	0.54 (0.42–0.71)	0.53 (0.40–0.71)	47 (29 to 60)
Vaccinated contact and unvaccinated case	3198	368	12	0.29 (0.26–0.33)	0.28 (0.24–0.33)	72 (67 to 76)
Vaccinated contact and case	5404	564	10	0.27 (0.24–0.29)	0.24 (0.21–0.29)	76 (71 to 79)
Contacts of index cases infected with the Alpha variant
Unvaccinated contact and case	2030	328	16	1	1	Reference
Unvaccinated contact and vaccinated case	23	5	22	1.35 (0.56–3.25)	1.22 (0.50–2.98)	–22 (–198 to 50)
Vaccinated contact and unvaccinated case	246	19	8	0.48 (0.30–0.76)	0.46 (0.28–0.77)	54 (23 to 72)
Vaccinated contact and case	32	0	0	NA	NA	NA
Contacts of index cases infected with the Delta variant
Unvaccinated contact and case	585	293	50	1	1	Reference
Unvaccinated contact and vaccinated case	101	28	28	0.55 (0.38–0.82)	0.55 (0.36–0.84)	45 (16 to 64)
Vaccinated contact and unvaccinated case	565	58	10	0.21 (0.16–0.27)	0.20 (0.14–0.28)	80 (72 to 86)
Vaccinated contact and case	1867	184	10	0.20 (0.16–0.24)	0.17 (0.12–0.24)	83 (76 to 88)
Contacts of index cases infected with variants other than Alpha and Delta
Unvaccinated contact and case	309	76	25	1	1	Reference
Unvaccinated contact and vaccinated case	3	0	0	NA	NA	NA
Vaccinated contact and unvaccinated case	80	8	10	0.41 (0.20–0.84)	0.57 (0.23–1.42)	43 (–42 to 77)
Vaccinated contact and case	22	4	18	0.74 (0.27–2.02)	0.87 (0.22–3.44)	13 (–244 to 78)

RR: relative risk; VE: vaccine effectiveness; NA: not available; contact means close contact; case means index case.

^aRelative risk adjusted by age groups (18–34, 35–49, 50–69 and ≥70 years), sex, major chronic condition of close contacts, and month of contact.

Vaccine effectiveness on SARS-CoV-2 transmission in the household

The secondary attack rate from unvaccinated index cases to their unvaccinated close household contacts was 45% on average. COVID-19 vaccination of the index case did not show a significant reduction in the infectiousness (13%, 95% CI −5% to 28%). A moderate preventive effect of transmission (61%, 95% CI 58–64%) was observed in vaccinated close contacts of unvaccinated index cases, while paradoxically, vaccination of both provided a lower reduction of the transmissibility (52%, 95% CI 47–56%; p < 0.001). This vaccine effect in reducing transmission was lower in close contacts older than 60 years (18%, 95% CI −6% to 36%) compared to those aged 18–59 years (58%, 95% CI 53–62%), as well as in close contacts without major chronic conditions (56%, 95% CI 50–61%) compare to those with any of these conditions (40%, 95% CI 28–51%). The paradoxical lower reduction of transmissibility when both, the close contact and the index case, were vaccinated, compared with only vaccinated close contact was mainly observed in contacts older than 60 years (18% vs 56%, p < 0.001) (Fig. 2 and Supplementary Tables S4-S6).

The secondary attack rates were higher in close contacts of symptomatic index cases than in those contacts of asymptomatic index cases, and the effect of the close contact vaccination was higher when the index case was symptomatic (62% vs 49%). The secondary attack rates in close household contacts remained unchanged from the Alpha variant period (April-June) to the Delta variant period (July-November). As compared to vaccinated close contacts of unvaccinated index cases, the vaccination effectiveness in reducing transmission was lower when both were immunized, overall (52% vs 61%, p < 0.001), for symptomatic index cases (53% vs 62%, p < 0.001), and in the Delta variant period (51% vs 62%, p = 0.026) ( Table 3 and Supplementary Table S5).

Table 3.

Effectiveness of COVID-19 vaccination on the SARS-CoV-2 transmission among close household contacts.

	Total nº	Positive nº	%	Crude RR (95% CI)	Adjusted RR (95% CI)a	Adjusted VE, % (95% CI)a
Contacts of symptomatic index case
Unvaccinated contact and case	5372	2641	49	1	1	Reference
Unvaccinated contact and vaccinated case	218	101	46	0.94 (0.77–1.15)	0.84 (0.69–1.04)	16 (−4 to 31)
Vaccinated contact and unvaccinated case	6026	1185	20	0.40 (0.37–0.43)	0.38 (0.35–0.41)	62 (59 to 65)
Vaccinated contact and case	3076	1024	33	0.68 (0.63–0.73)	0.47 (0.42–0.52)	53 (48 to 58)
Contacts of asymptomatic index case
Unvaccinated contact and case	1089	288	26	1	1	Reference
Unvaccinated contact and vaccinated case	47	16	34	1.29 (0.78–2.13)	1.11 (0.66–1.87)	–11 (–87 to 34)
Vaccinated contact and unvaccinated case	1490	238	16	0.60 (0.51–0.72)	0.51 (0.40–0.66)	49 (34 to 60)
Vaccinated contact and case	350	63	18	0.68 (0.52–0.89)	0.48 (0.34–0.69)	52 (31 to 66)
Contacts with major chronic condition
Unvaccinated contact and case	1537	689	45	1	1	Reference
Unvaccinated contact and vaccinated case	64	25	39	0.87 (0.59–1.30)	0.79 (0.52–1.19)	21 (–19 to 48)
Vaccinated contact and unvaccinated case	2076	370	18	0.40 (0.35–0.45)	0.40 (0.34–0.48)	60 (52 to 66)
Vaccinated contact and case	1194	425	36	0.79 (0.70–0.90)	0.60 (0.49–0.72)	40 (28 to 51)
Contacts without major chronic condition
Unvaccinated contact and case	4929	2240	46	1	1	Reference
Unvaccinated contact and vaccinated case	201	92	46	1.01 (0.82–1.24)	0.90 (0.72–1.11)	10 (–11 to 28)
Vaccinated contact and unvaccinated case	5430	1053	19	0.43 (0.40–0.46)	0.39 (0.35–0.43)	61 (57 to 65)
Vaccinated contact and case	2232	662	30	0.65 (0.60–0.71)	0.44 (0.39–0.50)	56 (50 to 61)
April to June 2021
Unvaccinated contact and case	4483	1999	45	1	1	Reference
Unvaccinated contact and vaccinated case	30	10	33	0.75 (0.40–1.39)	0.76 (0.41–1.41)	24 (–41 to 59)
Vaccinated contact and unvaccinated case	503	88	18	0.39 (0.32–0.49)	0.38 (0.30–0.47)	62 (53 to 70)
Vaccinated contact and case	32	6	19	0.42 (0.19–0.94)	0.37 (0.16–0.82)	63 (18 to 84)
July to November 2021
Unvaccinated contact and case	1978	930	47	1	1	Reference
Unvaccinated contact and vaccinated case	235	107	46	0.97 (0.79–1.18)	0.87 (0.71–1.06)	13 (–6 to 29)
Vaccinated contact and unvaccinated case	7013	1335	19	0.41 (0.37–0.44)	0.38 (0.35–0.43)	62 (57 to 65)
Vaccinated contact and case	3394	1071	32	0.68 (0.62–0.74)	0.49 (0.44–0.55)	51 (45 to 56)
Contacts of index cases infected with the Alpha variant
Unvaccinated contact and case	1760	739	42	1	1	Reference
Unvaccinated contact and vaccinated case	16	10	63	1.49 (0.80–2.78)	1.55 (0.83–2.91)	–55 (–191 to 17)
Vaccinated contact and unvaccinated case	337	39	12	0.28 (0.20–0.38)	0.27 (0.19–0.38)	73 (62 to 81)
Vaccinated contact and case	25	6	24	0.57 (0.26–1.28)	0.55 (0.24–1.26)	45 (−26 to 76)
Contacts of index cases infected with the Delta variant
Unvaccinated contact and case	352	182	52	1	1	Reference
Unvaccinated contact and vaccinated case	73	32	44	0.85 (0.58–1.23)	0.81 (0.55–1.18)	19 (–18 to 45)
Vaccinated contact and unvaccinated case	1165	225	19	0.37 (0.31–0.45)	0.37 (0.30–0.47)	63 (53 to 70)
Vaccinated contact and case	1042	338	32	0.63 (0.52–0.75)	0.46 (0.37–0.58)	54 (42 to 63)
Contacts of index cases infected with variants other than Alpha and Delta
Unvaccinated contact and case	178	77	43	1	1	Reference
Unvaccinated contact and vaccinated case	2	1	50	1.16 (0.16–8.31)	1.01 (0.14–7.39)	–1 (–639 to 86)
Vaccinated contact and unvaccinated case	145	20	14	0.32 (0.20–0.52)	0.47 (0.25–0.92)	53 (8 to 75)
Vaccinated contact and case	14	3	21	0.50 (0.16–1.57)	0.76 (0.22–2.68)	24 (–168 to 78)

RR: relative risk; VE: vaccine effectiveness; contact means close contact; case means index case.

^aRelative risk adjusted by age groups (18–34, 35–49, 50–69 and ≥70 years), sex, major chronic condition of close contacts, and month of contact.

Discussion

The results demonstrate that COVID-19 vaccination significantly reduces the susceptibility of persons exposed to SARS-CoV-2, and that infected persons who had previously been vaccinated against COVID-19 are less infectious for their close contacts. However, vaccination of both did not show complete synergy and these effects were lower in older people.

In the absence of COVID-19 vaccination, we found a risk of transmission from an infected person to their close social contacts of 26% on average. COVID-19 vaccination reduced the susceptibility of close contact by 69%, and the previously vaccinated infected persons were 44% less infectious for their contacts. Vaccination of the index case and close contact reduced the risk of SARS-CoV-2 transmission in social settings by 74%, slightly less than expected assuming a complete synergy of both vaccination effects.

Although SARS-CoV-2 transmission could happen from asymptomatic index cases, symptomatic cases were more infectious for their close contacts, as previous studies have observed [20], [25], [26]. In the present study, compared to the Alpha variant dominance period, the secondary attack rates within each vaccination category of the index case and contact were higher in the Delta variant period. A consistent finding was seen in the analysis of close contacts of index cases with the L452R mutation characteristic of the Delta variant [27]. Nevertheless, point estimates of effectiveness in preventing transmission by COVID-19 vaccination of the index case or close contact were similar or higher in the Delta variant period and when the Delta variant was detected in the index case, which contrasts with the results of other authors that have described lower vaccine effectiveness against the Delta variant [28]. However, these estimates are consistent with the more rapid decline in viral load observed in vaccinated cases infected by the Delta variant [29].

Compared with social settings, in the household setting, the risk of transmission between unvaccinated persons was higher and the vaccination effects in reducing infectiousness and transmission were lower. The more intense and repeated exposures in close household contacts may result in a higher risk of transmission and a reduced effect of preventive measures [5], [20]. In older people cohabiting, a paradoxical lower preventive effect of the close contact vaccination was observed when the index case was also vaccinated. The lack of a plausible biological explanation suggests that this lower protective effect may be due to an increased risk assumption and relaxation in non-pharmacological preventive measures when the index case and the close household contact were both vaccinated.

Early studies on the effectiveness of the COVID-19 vaccines reported a very low risk of infection in vaccinated people and a low risk of transmission from vaccinated cases [6], [10]. As a result of these findings, the protocols for COVID-19 control proposed relaxer recommendations for vaccinated than unvaccinated close contacts [21], and the population could misunderstand that other preventive measures were no longer necessary among fully vaccinated persons. Since general mandatory measures, such as facemask use or social distancing were maintained in social settings, the relaxation mainly affected household contacts. Further studies have evidenced that the vaccination effect in preventing transmission was only partial and that additional preventive measures are necessary for efficient SARS-CoV-2 transmission control [9], [30].

The mentioned paradoxical decrease in the preventive effect was mainly observed in older close household contacts. In households, compliance to preventive measures is difficult to maintain during continuous exposure and the vaccination status of the other cohabitants is usually well known, so fully vaccinated cohabitants could easily assume relaxation in preventive measures. The consequences of this lower protection are worse in older people due to the lower and less lasting effect of the vaccines [7].

The possible relaxation of preventive measures following vaccination may introduce a bias in studies of COVID-19 vaccine effectiveness in preventing infections; therefore, the results of studies that do not achieve good control of this bias may underestimate vaccine effectiveness. Furthermore, the subsequent changes in preventive behaviors could hide part of the impact of vaccination on SARS-CoV-2 transmission.

The present results are consistent with previous studies showing that the current vaccination programs are not able to achieve complete control of SARS-CoV-2 transmission, especially in ageing populations [7]. In these cases, complementary preventive measures may be necessary, as social distancing and use of facemasks [8].

The strengths of this study are the prospective open cohort design of close contacts of confirmed cases of COVID-19. The recruitment was based on a contact-tracing protocol, which was the same for all the population and remained unchanged during the study period. As two cohorts were studied, the comparison of their results provides complementary views of the effect of vaccination on SARS-CoV-2 transmission under different conditions. Variables were obtained before the outcome was known. Close contacts were classified according to the test result.

This study may have limitations. Close contacts that did not complete the testing protocol (3.3%; 2027/62,355) were excluded from the analysis; therefore potential selection bias due to loss of follow up would be small. Vaccine effectiveness estimates may be conditioned by the time since vaccination and the circulating SARS-CoV-2 strains during the study period; therefore, these results may not be valid for earlier or later periods [31]. Vaccination coverage varied with age, major chronic conditions, and month, but analyses were adjusted for these variables. Symptomatic patients with positive antigen test were included; however, a previous study found no interference in estimates [7]. People younger than 18 years were excluded because they were not the target population for vaccination during almost the entire study period. The effect of vaccination could be affected by the possible relaxation in preventive measures after vaccination. Although the study population was composed of close contacts of infected cases, we cannot rule out some differential changes in the frequency and intensity of the contact after vaccination. These changes might bias the vaccine effectiveness estimates. The paradoxical results found in old household contacts, but not in other groups, strongly suggest the effect of behavioral changes after vaccination. The effect of partial vaccination on transmission was based on an insufficient number of cases to achieve conclusions. During the study period booster doses were not used and the Omicron variant was not detected; therefore, we cannot conclude on the vaccination effect in these contexts [32]. Other factors such as the time since the last vaccine dose and the vaccine brand may modify the vaccine effect and were not considered in the present study. SARS-CoV-2 sequencing was only available in a part of the index cases; however, the results defined periods of dominance of each variant.

In conclusion, the results support that COVID-19 vaccination significantly reduces the susceptibility of persons exposed to the SARS-CoV-2 and that infected persons vaccinated against COVID-19 are less infectious for their contacts. However, these effects are insufficient for total control of SARS-CoV-2 transmission, since fully vaccinated people may be infected and transmit the infection for their close contacts. Furthermore, these results suggest that relaxation of preventive behaviors may counteract part of the vaccine effect on transmission, especially in older people in the household. Therefore, non-pharmaceutical preventive measures remain relevant around highly vulnerable people even among fully vaccinated individuals, as long as there is evidence of SARS-CoV-2 circulation.

Funding

This study was supported by the Horizon 2020 program of the European Commission (I-MOVE-COVID-19, grant agreement No 101003673), and by the Carlos III Institute of Health with the European Regional Development Fund (COV20/00542, CM19/00154, INT21/00100, and CP22/00016). The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.

Conflict of interest

All authors have no potential conflict of interest.

Appendix A. Supplementary material

Supplementary material

.