Table 2.
Protection conferred by previous infection and hybrid immunity compared to immune-naive individuals
| Number of studies | Number of estimates | Month 1* | Month 2† | Month 3 | Month 4 | Month 6 | Month 9 | Month 12 | Month 15 | Percentage point change in protection, 3–6 months (95% CI)‡ | Percentage point change in protection, 3–12 months (95% CI)‡ | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Previous infection | ||||||||||||
| Hospital admission or severe disease | 6 | 16 | NA | 83·2% (72·1 to 90·5) | 82·5% (71·8 to 89·7) | 81·7% (71·4 to 88·9) | 80·1% (70·3 to 87·2) | 77·5% (67·5 to 85·1) | 74·6% (63·1 to 83·5) | 71·6% (57·1 to 82·6) | −2·4 (−5·1 to 4·7) | −7·8 (−20·9 to 12·1) |
| Any infection§ | 10 | 64 | NA | 69·5% (57·6 to 79·2) | 65·2% (52·9 to 75·9) | 60·7% (48 to 72·1) | 51·2% (38·6 to 63·7) | 37·0% (26 to 49·6) | 24·7% (16·4 to 35·5) | 15·5% (9·9 to 23·6) | −14·0 (−12·0 to −18·2) | −40·5 (−33·9 to −51·9) |
| Hybrid immunity (primary series vaccination) | ||||||||||||
| Hospital admission or severe disease | 5 | 23 | 95·7% (88·0 to 98·5) | 95·9% (88·5 to 98·6) | 96·0% (89·0 to 98·6) | 96·2% (89·4 to 98·7) | 96·5% (90·2 to 98·8) | 97·0% (90·9 to 99) | 97·4% (91·4 to 99·2)¶ | NA | 0·50 (−2·2 to 2·1) | 1·3 (−4·3 to 7·4) |
| Any infection | 7 | 55 | 74·1% (64·8 to 81·6) | 71·6% (61·9 to 79·6) | 69·0% (58·9 to 77·5) | 66·2% (55·8 to 75·3) | 60·4% (49·6 to 70·3) | 51·1% (40·2 to 61·9) | 41·8% (31·5 to 52·8)¶ | NA | −8·6 (−1·7 to −17·2) | −27·2 (−6·4 to −53·2) |
| Hybrid immunity (first booster vaccination) | ||||||||||||
| Hospital admission or severe disease | 4 | 17 | 98·0% (92·9 to 99·5) | 97·6% (91·6 to 99·4) | 97·2% (90·0 to 99·3) | 96·7% (87·9 to 99·1) | 95·3% (81·9 to 98·9)¶ | NA | NA | NA | −1·8 (−10·3 to 0·77) | NA |
| Any infection | 6 | 24 | 80·1% (72·5 to 86) | 74·8 (66·0 to 81·9) | 68·6% (58·8 to 76·9) | 61·6% (51·2 to 71·1) | 46·5% (36·0 to 57·3)¶ | NA | NA | NA | −22·0 (−4·3 to −38·8) | NA |
This table displays the point estimates and 95% CIs of protection shown in figure 2. This analysis used a log-odds meta-regression model. NA=not available (ie, insufficient data for model extrapolation).
Month 1 data were for individuals with hybrid immunity whose last immunological challenge was vaccination and thus were eligible for reinfection within a shorter timeframe than people who most recently had previous infection (2 months minimum for probable reinfection).
Month 2 data represent the minimum time period for reinfection among individuals with previous infection (ie, possible reinfection).
95% CIs calculated with the bootstrap method. Percentage point changes over time are reported from 3 months as this represents probable and confirmed reinfections.
Any infections comprised mild infections, symptomatic infections, and asymptomatic infections.
Model predictions beyond the range of the available data. Previous infection data are available for predictions for 2–16 months; hybrid immunity data were available for predictions for 1–11 months. Data were extrapolated to a maximum of 3 months beyond the final follow-up date.