Temporal changes in the positivity rate of common enteric viruses among paediatric admissions in coastal Kenya, in the period spanning the COVID-19 pandemic, 2019-2022

20

21 The non-pharmaceuMcal intervenMons (NPIs) implemented to curb the spread of SARS-CoV-2 22 early in the COVID -19 pandemic years, disrupted the acMvity of other respiratory viruses. 23 There is limited data from low-and-middle income countries (LMICs ) to determine whether 24 COVID-19 NPIs also impacted the epidemiology of enteric viruses. We invesMgated the 25 changes in infecMon pacerns of common enteric viruses among hospitalised children who 26 presented with diarrhoea to a referral hospital in coastal Kenya, in the period spanning the 27 COVID-19 pandemic. 28 29

30 A total of 870 stool samples from children under 13 years of age admiced to Kilifi County 31 Hospital between January 2019, and December 2022 were screened for rotavirus group A 32 (RVA), norovirus genogroup II (GII), astrovirus, sapovirus, and adenovirus type F40/41 using 33 real-Mme reverse-transcripMon polymerase chain reacMon. The proporMons posiMve across the 34 four years were compared using the chi-squared test staMsMc. 35 36

37 One or more of the five virus targets were detected in 282 (32.4%) cases. A reducMon in the 38 posiMvity rate of RVA cases was observed from 2019 (12.1%, 95% confidence interval (CI) 8.7% 39 -16.2%) to 2020 (1.7%, 95% CI 0.2% – 6.0%; p < 0.001). However, in the 2022, RVA posiMvity 40 rate rebounded to 23.5% (95% CI 18.2% - 29.4%). For norovirus GII, the posiMvity rate 41 fluctuated over the four years with its highest posiMvity rate observed in 2020 (16.2%; 95% 42 C.I, 10.0% – 24.1%). No astrovirus cases were detected in 2020 and 2021 , but the posiMvity 43 rate in 2022 was similar to that in 2019 (3.1% (95% CI 1.5% - 5.7%) vs 3.3% (95% CI 1.4% – 44 6.5%)). A higher case fatality rate was observed in 2021 (9.0%) compared to the 2019 (3.2%), 45 2020 (6.8%) and 2022 (2.1%) (p <0.001). 46 47

48 Our study finds that in 2020 the transmission of common enteric viruses, especially RVA and 49 astrovirus, in Kilifi Kenya may have been disrupted due to the COVID-19 NPIs. Aker 2020, local 50 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 24, 2023. ; https://doi.org/10.1101/2023.07.24.23293059doi: medRxiv preprint 3 enteric virus transmission pacerns appeared to return to pre-pandemic levels coinciding with 51 the removal of most of the government COVID-19 NPIs. 52 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 24, 2023. ; https://doi.org/10.1101/2023.07.24.23293059doi: medRxiv preprint 4 Introduc-on 53 Although water sanitaMon and hygiene (WASH) programmes and new vaccine introducMons 54 have resulted in significant reducMons of paediatric diarrhoea morbidity and mortality 55 globally, virus-associated diarrhoea is sMll a major cause of hospital admissions in several low 56 and middle-income semngs 1. In 2019, approximately 300,000 deaths were recorded globally 57 in children below 14 years of age due to rotavirus group A (RVA), norovirus GI and G II, and 58 adenovirus 40/41 infecMons 2. 59 60 Following the emergence of COVID-19, a number of reports have indicated perturbaMons in 61 in seasonality, prevalence and incidence of common enteric viruses associated with diarrhoea 62 disease. For instance, in France 3, Poland 4, China 5,6 and USA 7, the prevalence of RVA during 63 the year 2020 was lower compared to 2018 and 2019. However, in 2021 there was a surge of 64 RVA cases in these countries. Like RVA, a decrease in cases of norovirus GII, sapovirus, 65 adenovirus 40/41 and astrovirus was reported in 2020 in Spain 8 and Korea 9 compared to 66 previous years. However, this decrease in virus detecMon has been followed by remarkable 67 outbreaks in 2021 5,10. Sporadic outbreaks of norovirus have also been reported in China in 68 September 2020 and in the USA, where a total of 992 norovirus outbreaks were reported 69 between August 2021 and July 2022 5,11,12. The decline of the detecMon rates of enteric viruses 70 in the early COVID-19 pandemic phase has been postulated to be a result of the stringent non-71 pharmaceuMcal intervenMons (NPIs) that were implemented to abrogate the pandemic 13. 72 Some of the measures included those that may impact enteric pathogen transmission such as 73 frequent hand washing, increased hygiene, social distancing, closure of restaurants and 74 restricted movement either locally or internaMonally 13. 75 76 In coastal Kenya, the prevalence of enteric viruses over the past decade has been monitored 77 and no significant change in the prevalence has been detected in all the viruses except for 78 sapovirus (7.6% vs 4.0%, p value <0.05) pre-post rotavirus vaccine introducMon in July 2014 14. 79 RVA posiMvity in hospital admissions decreased significantly only among ELISA detected cases 80 but not RT-PCR detected cases 15. ConMnuous monitoring of these enteric viruses that 81 commonly cause diarrhoea in childhood is key in providing insights on their epidemiology for 82 disease management and informing public health policy. In this study, we aimed to describe 83 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 24, 2023. ; https://doi.org/10.1101/2023.07.24.23293059doi: medRxiv preprint 5 the epidemiological pacerns of common enteric viruses associated with diarrhoea between 84 January 2019 and December 2022, during a period spanning the COVID-19 pandemic. 85 86

87 Study site and popula-on 88 This study was undertaken as part of our rouMne surveillance of RVA in Kilifi County Hospital 89 (KCH), Kenya 14,16,17. To be recruited, a parMcipant had to saMsfy the following criteria: (a) 90 admiced with diarrhoea as one of their illness symptom(s), (b) aged < 13-year-old, (c) consent 91 given from a parent or guardian to be in the study 14,16,17 The surveillance started in 2009 and 92 has conMnued to date (2023). In this analysis we focused on parMcipants recruited between 93 1st January 2019 and 31st December 2022. 94 95 Laboratory methods. 96 Molecular tes-ng for common enteric viruses 97 Total Nucleic Acid (TNA) Extrac3on 98 TNA was extracted from 0.2 grams of stool (or 200ul if liquid) using the QIAamp Fast DNA Stool 99 Mini kit (Qiagen, Manchester, UK) and eluted in 200ul of eluMon buffer as previously described 100 14,16. 101 102 Virus (RT)-PCR Screening 103 The extracted TNA was combined with the TaqMan Fast Virus 1-step master mix and virus 104 specific primers (supplementary table 1) for each of the five viruses 16,18 .and processed on a 105 real-Mme QuanMstudio 5 -flex instrument. The reacMon mix comprised 2.5µl of the TaqMan 106 master mix, 1.2µl of the primer-probe mix, 3.8µl of nuclease free water and 5µl of TNA. The 107 thermocycling condiMons were as follows; 95°C for 20 sec and 35 cycles of 94°C for 15 sec and 108 60°C for 30 sec. A cycle threshold cut-off of < 35.0 was applied to determine virus posiMve 109 samples for all targets screened. 110 111 RVA genotyping 112 TNA from RVA posiMves were amplified using segment specific primers, sequenced on the 113 Illumina Miseq plarorm as previously described 19. Genomes were assembled from the short 114 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 24, 2023. ; https://doi.org/10.1101/2023.07.24.23293059doi: medRxiv preprint 6 read data using a de novo assembly approach as previously described 19. RVA genotypes were 115 assigned using either BLAST or an online RVA genotyping tool 20. 116 117 Sta-s-cal Analysis 118 All staMsMcal analysis was undertaken using R version 4.1.1 (2021 -08-10). The level of 119 government intervenMon was summarised using the Oxford Stringency index (SI), a composite 120 measure based on nine response indicators including school closures, workplace closures, and 121 travel bans, rescaled to a value from 0 to 100 (100 = strictest) 21. Local stringency measures 122 have been highlighted elsewhere 22 and summarized in supplementary table 2. 123 124 The virus posiMvity rate during each year was calculated as the proporMon of samples that 125 tested posiMve for the given virus given the total number of samples tested in the defined 126 year. The data from 2019 has been previously published elsewhere and formed a reference 127 base of the situaMon before COVID-19 16. Comparisons across different years and groups were 128 done using the chi-squared test staMsMc. Kruskal Wallis and Wilcoxon rank-sum tests were 129 used to compare the distribuMon of conMnuous variables. Disease severity was esMmated 130 using the Vesikari Clinical Severity Scoring System Manual as previously described 16,23.131 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 24, 2023. ; https://doi.org/10.1101/2023.07.24.23293059doi: medRxiv preprint 7

132 Basic demographic characteris-cs 133 Between January 2019 and December 2022, 1,613 paMents aged under 13 years presented 134 with diarrhoea as one of their illness symptoms at KCH. Of these, 870 (54.0%) consented 135 enrolment into the study, gave a stool sample, and were included in this analysis. The reasons 136 for missed sample collecMon in the study were: consent refusal (n=344), other (n=133), death 137 (n=68), discharged before sample collecMon(n=35) and transferred before sample collecMon 138 (n=3). 139 140 All the 870 stool samples were screened for the five enteric viruses namely, RVA, norovirus 141 GII, astrovirus, sapovirus and adenovirus F40/41. The majority of the recruited paMents were 142 in their first year of life (n=371, 42.6%) and all suffered moderate-to-severe diarrhoeal disease 143 (Table 1). The characterisMcs of the observed cases across the four years in terms of gender 144 and age were similar (p value > 0.05). However, fatal outcome appeared more likely to occur 145 in 2021 (9.0%) compared to 2019 (3.2%), 2020 (6.8%) and 2022 (2.1%) (p value < 0.001, Table 146 1). Less severe disease was also reported in 2020 compared to the other three years, Table 1.147 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 24, 2023. ; https://doi.org/10.1101/2023.07.24.23293059doi: medRxiv preprint 8 Table 1: Demographic characterisMcs of children under 13 years admiced to Kilifi County 148 Hospital, Kenya between January 2019 and December 2022 149 2019 (n=314) * 2020 (n=117) 2021 (n=201) 2022 (n=238) Total (n=870) P value Sex 0.586 Female 137 (43.6%) 45 (38.5%) 81 (40.3%) 107 (45.0%) 370 (42.5%) Male 177 (56.4%) 72 (61.5%) 120 (59.7%) 131 (55.0%) 500 (57.5%) Median age in months (Interquartile range) 14.0 (8.0 – 25.0) 13.7 (8.1 – 27.0) 13.0 (7.5 – 23.4) 13.2 (8.5 – 21.3) 13.7 (8.1 – 23.0) Age strata (months) 0.252 <12 128 (40.8%) 51 (43.6%) 89 (44.3%) 103 (43.3%) 371 (42.6%) 12 -23 102 (32.5%) 36 (30.8%) 64 (31.8%) 86 (36.1%) 288 (33.1%) 24 -59 56 (17.8%) 13 (11.1%) 33 (16.4%) 35 (14.7%) 137 (15.7%) >60 28 (8.9%) 17 (14.5%) 15 (7.5%) 14 (5.9%) 74 (8.5%) Outcome <0.001 Alive 276 (87.9%) 109 (93.2%) 183 (91.0%) 231 (97.1%) 799 (91.8%) Dead 10 (3.2%) 8 (6.8%) 18 (9.0%) 5 (2.1%) 41 (4.7%) Data missing 28 (8.9%) 0 (0.0%) 0 (0.0%) 2 (0.8%) 30 (3.4%) Disease severity 0.01 Moderate 90 (28.7%) 47 (40.2%) 45 (22.4%) 65 (27.3%) 247 (28.4%) Severe 224 (71.3%) 70 (59.8%) 156 (77.6%) 173 (72.7%) 623 (71.6%) *Pre-pandemic data has been reported in an earlier manuscript 16 150 Trends in diarrhoeal cases in the context of the COVID-19 pandemic 151 Aker the iniMal detecMon of the first COVID-19 case in Kenya on 12 th March 2020 24, the 152 government implemented a range of NPIs to curb the pandemic (summarised in Figure 1a 153 using the Oxford Stringency index and local restricMons listed in supplementary table 2). The 154 trend of monthly recorded diarrhoea cases recruited into our surveillance between January 155 2019 and December 2022 is shown in Figure 1b. The highest monthly peak in cases was 156 recorded in 2019 before the COVID-19 pandemic. The lowest number of diarrhoea cases was 157 in the 2020, but a gradual rebound was observed in 2021 and 2022. 158 159 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 24, 2023. ; https://doi.org/10.1101/2023.07.24.23293059doi: medRxiv preprint 9 160 Figure 1: A) COVID-19 stringency index quanMfying the government NPI measures aimed to 161 curb the spread of SARS-CoV-2 (source: hcps://ourworldindata.org/covid-stringency-index). 162 B) Temporal trends of monthly diarrhoea and virus cases between January 2019 and 163 December 2022. The blue and red line graphs show the total eligible and recruited diarrhoea 164 cases respecMvely, while the bar graphs show the total virus posiMve cases per month. 165 166 Single virus infec-ons and coinfec-ons 167 At least one of the five virus targets were detected in 282 (32.4%) cases. The proporMon of 168 samples posiMve for the five viruses we tested in the stool samples for the different years is 169 summarised in table 1. The posiMvity rate of RVA deeped in 2020, at 1.7% (95% C.I, 0.2% – 170 6.0%) compared to 2019 (12.1% (95% C.I, 8.7%-14.9%) and gradually rose in 2021 (16.9% (95% 171 C.I, 12.0% - 22.8%)) and 2022 (23.5% ( 95% C.I, 18.2% - 29.4%)) and the differences were 172 staMsMcally significant (p value < 0.001) (Table 2). 173 174 The norovirus GII posiMvity rate fluctuated over the four years and the difference was 175 staMsMcally significant (p value 0.04) (Table 2). The highest posiMvity rate for norovirus GII was 176 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 24, 2023. ; https://doi.org/10.1101/2023.07.24.23293059doi: medRxiv preprint 10 16.2% (95% C.I, 10.0% – 24.1%) in 2020 and lowest at 6.4% (95% C.I, 3.4% – 10.8%) in 2021 177 (Table 2). No astrovirus cases were detected in the 2020 and 2021 , but the cases in 2022 178 (3.3%) with a similar posiMvity rate to what was reported in 2019 (3.1%, Table 2). The posiMvity 179 rates for sapovirus and adenovirus type F40/41 did not change across the three phases (p 180 value > 0.05, χ2). 181 182 Table 2: Comparison of the detecMon rates of five common enteric virus between January 183 2019 and December 2022 184 Total (n=870) 2019 (n=314) 2020 (n=117) 2021 (n=201) 2022 (n=238) P value Cases Proportion (95% CI) Cases Proportion (95% CI) Cases Proportion (95% CI) Cases Proportio n (95% CI) Virus Rotavirus 130 (14.9%) 38 12.1 (8.7 - 16.2) 2 1.7 (0.2 – 6.0) 34 16.9 (12.0 – 22.8) 56 23.5 (18.2 – 29.4) <0.00 1 Norovirus GII 89 (10.2%) 34 10.8 (7.6 – 14.8) 19 16.2 (10.0 – 24.1) 13 6.4 (3.4 - 10.8) 23 9.6 (6.2 – 14.1) 0.04 Astrovirus 18 (2.1%) 10 3.1 (1.5 – 5.7) 0 - 0 - 8 3.3 (1.4 – 6.5) 0.01 Sapovirus 33 (3.8%) 11 3.5 (1.7 – 6.1) 3 2.5 (0.5 – 7.3) 7 3.4 (1.4 – 7.0) 12 5.0 (2.6 – 8.6) 0.65 Adenovirus F40/41 30 (3.4%) 9 2.8 (1.3– 5.3) 8 6.8 (2.9 - 13.0) 6 2.9 (1.1 – 6.3) 7 2.9 (1.1 – 5.9) 0.19 185 Between 2019 and 2022, 13 samples had a coinfecMon of two or more viruses of the screened 186 viruses. The most common coinfecMons were RVA and norovirus GII (n=4), adenovirus 40/41 187 and astrovirus (n=3), and adenovirus 40/41 and sapovirus (n=3) (Supplementary table 3). 188 189 Monthly virus trends 190 In all the years, peak RVA cases were observed in the month of August, except in 2020 where 191 only two RVA cases were detected, Figure 2. The peak months for norovirus GII varied across 192 the different years. Less than five cases were reported over the four years for adenovirus 193 40/41, sapovirus and astrovirus in each month. In 2020 and 2021 , no astrovirus cases were 194 detected but re-emerged in 2022 (Figure 2). 195 196 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 24, 2023. ; https://doi.org/10.1101/2023.07.24.23293059doi: medRxiv preprint 11 197 198 199 Figure 2: Monthly temporal distribuMon of common enteric virus cases in children under 13 200 years admiced to KCH with diarrhoea between January 2019 and December 2022. The black 201 trendline shows the monthly diarrhoeal cases over Mme and the verMcal doced line represents 202 when the first COVID-19 case was detected in Kenya. 203 204 205 RVA genotypes and vaccina-on status of posi-ve cases 206 Of the 130 RVA posiMve detected during the study, 87 (66.9%) and 70 (53.8%) successively 207 sequenced in the VP7 and VP4 segment, respecMvely. The years 2019, 2020 and 2021 were 208 predominated by the G3P[8] genotype (n=48, 80.0%). In 2022, we observed replacement of 209 the G3P[8] genotype with mulMple genotypes: G2P[4] (n=2), G9P[8] (n=10), and G9P[4] (n=4). 210 To note, in the 2022 there was some incomplete genotyping due to failed sequencing in the 211 VP4 and VP7 segments G2P[x] (n=5) and GxP[8] (n=8), Among the 130 RVA posiMve cases, 88 212 0 5 10 15 20 25 30 35 0 5 10 15 20 25 30 35 40 45 50 55 Virus cases Rotavirus A 0 5 10 15 20 25 30 35 0 5 10 15 20 25 30 35 40 45 50 55 Virus cases Norovirus GII 0 5 10 15 20 25 30 35 0 5 10 15 20 25 30 35 40 45 50 55 Jan−2019 Apr−2019 Jul−2019 Oct−2019 Jan−2020 Apr−2020 Jul−2020 Oct−2020 Jan−2021 Apr−2021 Jul−2021 Oct−2021 Jan−2022 Apr−2022 Jul−2022 Oct−2022 Jan−2023 Period (2019−2022) Viirus cases Adenovirus F40/41 0 5 10 15 20 25 30 35 0 5 10 15 20 25 30 35 40 45 50 55 Virus cases Sapovirus 0 5 10 15 20 25 30 35 0 5 10 15 20 25 30 35 40 45 50 55 Jan−2019 Apr−2019 Jul−2019 Oct−2019 Jan−2020 Apr−2020 Jul−2020 Oct−2020 Jan−2021 Apr−2021 Jul−2021 Oct−2021 Jan−2022 Apr−2022 Jul−2022 Oct−2022 Jan−2023 Period (2019−2022) Virus cases Astrovirus . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 24, 2023. ; https://doi.org/10.1101/2023.07.24.23293059doi: medRxiv preprint 12 (67.7%) had received at least one dose of the Rotarix® vaccine two doses, 18 (12.9%) had not 213 received a vaccine and 24 (16.9%) had no vaccinaMon records. One parMcipant was RVA 214 posiMve five days aker receiving a Rotarix dose and their sample genotyped as G3P[8]. 215 216 217 Disease outcome 218 In the study, a total of 41 (4.7%) cases succumbed among the 870 that we analysed. Only nine 219 of these cases were posiMve for at least of the viruses we tested i.e., norovirus GII (n=4), 220 adenovirus type F40/41 (n=2), RVA (n=1), sapovirus (n=1) and a coinfecMon (RVA & sapovirus, 221 n=1) (Supplementary Figure 2). The other 32 cases had none of the five viruses detected. 222 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 24, 2023. ; https://doi.org/10.1101/2023.07.24.23293059doi: medRxiv preprint 13

223 In this coastal Kenya hospital-based study, we observed a decrease in paediatric diarrhoea 224 admissions in during the first year of the COVID-19 pandemic (2020) compared to the pre-225 pandemic year (2019). This decline may be acributed in part to be a result of the 226 implementaMon of NPIs such as restricMon of movement and increased hygiene pracMces that 227 may have led to reduce transmission of the enteric pathogens as observed elsewhere 8 and 228 the overall reduced access to the hospital occasioned by health worker strike between 229 December 2020. Peak diarrhoea cases are usually observed between June and July and in 230 2020 and 2021 , this period coincided with high stringency measures in the country. Aker 231 March 2021, targeted vaccinaMon campaigns started replacing NPIs and from August 2021, 232 the government dropped measures including closure of schools, curfews, lockdowns and 233 public gatherings and restricMons in public transport. 234 235 We report a significant decrease in RVA posiMvity rate in 2020 compared to 2019 followed by 236 a rebound in 2021 and 2022, a finding consistent with other studies elsewhere 4,6–8. Notably, 237 the posiMvity rate of RVA in 2022 was higher compared to all the previous years. Such a 238 resurgence of RVA acMvity was also observed in Hong Kong aker the first year of the 239 pandemic5. 240 241 The return and increase of RVA acMvity in the 2021 and 2022 can be acributed in part to the 242 relaxaMon of the COVID-19 NPIs or circulaMon of strains heterologous to the Rotarix vaccine 243 (e.g. G9P[8], G9P[4] and G2P[4]). Some of these strains have been noted to have limited cross-244 reacMvity with the Rotarix G1P[8] strain. Further, vaccinaMon delays and Rotarix® vaccine 245 stockouts that occurred in between June 2022 and January 2023 when a vaccine switch to 246 Rotavac was made in Kenya 25. The introducMon of G9 genotypes in Malaysia was associated 247 with an increase RVA prevalence 26 similar to what we observed in 2022 whereby the G3P[8] 248 were rapidly replaced by the G9P[8], G9P[4] and G2P[4] genotypes and subsequently there 249 was an increase in RVA cases. 250 251 In Kilifi, the prevalence of norovirus GII has been on the rise post -rotavirus vaccine 252 introducMon compared to the pre-vaccine period 14,16. InteresMngly, the posiMvity rate of 253 norovirus GII was highest in 2020 phase compared to other years. This implies that norovirus 254 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 24, 2023. ; https://doi.org/10.1101/2023.07.24.23293059doi: medRxiv preprint 14 acMvity was less impacted by the NPI measures which is known to be highly infecMous. An 255 increase in norovirus acMvity and mulMple sporadic outbreaks have reported in the USA and 256 China during the COVID-19 phase 5,11,12. 257 258 Notably, astrovirus was not detected in the 2020 and 2021 but re-emerged in the 2022 with a 259 posiMvity rate similar to what was observed in 2019. Astrovirus and sapovirus detecMon on 260 the Kenyan coast has been always characterised by very low prevalence (<5%) 14,16. 261 262 In China, enteric virus coinfecMons especially with RVA and norovirus was associated with 263 severe disease 27. Similarly in coastal Kenya, a coinfecMon with either RVA or/and norovirus 264 GII coincided with severe disease. However, its key to note that 71.6% of the parMcipants in 265 the s tudy presented with severe disease. In 2020, there was a significant difference in 266 mortality rate (9.0%) compared to the 2019 (3.2%), 2021 (6.8%) and 2022 (2.1) . We 267 hypothesize that the high mortality rate in the 2020 and partly 2021 may have been caused 268 by the challenges with access to the hospitals and health care services as efforts were geared 269 towards management of COVID-19 cases. 270 271 This study had several limitaMons. We did not analyse healthy controls from the same 272 populaMon to adjust the aeMological fracMon for asymptomaMc carriage in the populaMon. 273 With the small number of detecMons for some of the virus targets, it hard to confidently infer 274 seasonality without a larger study. There was substanMal data missingness e.g., on deaths and 275 almost half of the eligible parMcipants refused consent to be in the study . For a conclusive 276 inference of changes of the incidence of these viruses before during and aker the pandemic, 277 a populaMon-based study with a clear populaMon-based denominator is necessary. 278 279 In conclusion, our study observed a decrease in total diarrhoea admissions and enteric virus 280 acMvity during in 2020 and 2021. However, in 2022 an increase in RVA and astrovirus acMvity 281 is restored to pre-pandemic numbers. Therefore, conMnuous enteric virus surveillance is key 282 in understanding the temporal changes in posiMvity rate of these viruses to inform public 283 health policy. 284 285 286 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 24, 2023. ; https://doi.org/10.1101/2023.07.24.23293059doi: medRxiv preprint 15 Declara-ons 287 Ethics approval and consent to par-cipate. 288 The research protocol for the study was approved at Kenya Medical Research InsMtute 289 (KEMRI), by the ScienMfic and Ethics Review Unit (SSC#2861) in Nairobi, Kenya. 290

291 We are grateful to the study parMcipants who provided samples and members of the 292 pathogen epidemiology and omics group at KEMRI-Wellcome Trust Programme who did 293 sample collecMon and laboratory processing. This manuscript was wricen with the 294 permission of Director KEMRI CGMRC 295 Funding. 296 This study was funded by the Wellcome Trust (102975, 220985 and 226002/Z/22/Z). Dr 297 Charles AgoM was supported by the IniMaMve to Develop African Research Leaders (IDeAL) 298 through the DELTAS Africa IniMaMve [DEL-15-003]. The DELTAS Africa IniMaMve is an 299 independent funding scheme of the African Academy of Sciences (AAS)’s Alliance for 300 AcceleraMng Excellence in Science in Africa (AESA) and supported by the New Partnership for 301 Africa’s Development Planning and CoordinaMng Agency (NEPAD Agency). The views 302 expressed in this report are those of the authors and not necessarily those of AAS, NEPAD 303 Agency and The Wellcome. This research was funded in whole or in part by the Wellcome 304 Trust [102975, 220985 and 226002/Z/22/Z], For the purpose of Open Access, the author has 305 applied a CC-BY public copyright license to any author accepted manuscript version arising 306 from this submission. 307 Consent for publica-on 308 Yes. 309 Compe-ng interests 310 The authors declare no conflict of interest. 311 Availability of data and materials 312 All the data used can be accessed at the KWTRP Research repository via 313 hcps://doi.org/10.7910/DVN/EG6MEH. 314 Authors' contribu-ons 315 CAN and DJN sourced the study funding. CNA, DJN and AWL designed the study laboratory 316 assay. AWL, MM, TOM, CR, and GM did the laboratory experiments. NM and AWL managed 317 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 24, 2023. ; https://doi.org/10.1101/2023.07.24.23293059doi: medRxiv preprint 16 the study data and did the data analysis. AWL and CAN wrote the first manuscript drak. All 318 authors read, revised, and approved the final manuscript.319 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 24, 2023. ; https://doi.org/10.1101/2023.07.24.23293059doi: medRxiv preprint 17

320 1. Reiner RC, Wiens KE, Deshpande A, Baumann MM, Lindstedt PA, Blacker BF, et al. 321 Mapping geographical inequaliMes in childhood diarrhoeal morbidity and mortality in 322 low-income and middle-income countries, 2000–17: analysis for the Global Burden of 323 Disease Study 2017. Lancet [Internet]. 2020 Jun;395(10239):1779–801. Available 324 from: hcps://linkinghub.elsevier.com/retrieve/pii/S0140673620301148 325 2. InsMtute for Health Metrics and EvaluaMon I. Default results are global all-cause 326 deaths and DALYs for 2019 with trends since 1990 [Internet]. 2022 [cited 2022 Sep 327 27]. Available from: hcps://vizhub.healthdata.org/gbd-results/ 328 3. Cohen PR, Rybak A, Werner A, Béchet S, Desandes R, Hassid F, et al. Trends in 329 pediatric ambulatory community acquired infecMons before and during COVID-19 330 pandemic: A prospecMve mulMcentric surveillance study in France. Lancet Reg Heal - 331 Eur. 2022;22:1–10. 332 4. Toczylowski K, Jackowska K, Lewandowski D, Kurylonek S, Waszkiewicz-Stojda M, 333 Sulik A. Rotavirus gastroenteriMs in children hospitalized in northeastern Poland in 334 2006–2020: Severity, seasonal trends, and impact of immunizaMon. Int J Infect Dis 335 [Internet]. 2021;108:550–6. Available from: 336 hcps://doi.org/10.1016/j.ijid.2021.05.070 337 5. Chan MCW. Return of Norovirus and Rotavirus AcMvity in Winter 2020‒21 in City with 338 Strict COVID-19 Control Strategy, China. Emerg Infect Dis [Internet]. 2022 339 Mar;28(3):713–6. Available from: hcps://wwwnc.cdc.gov/eid/arMcle/28/3/21-340 2117_arMcle.htm 341 6. Fang C, Zhou Z, Li J, Zhou M. Incident changes of rotavirus enteriMs among children 342 during the coronavirus disease-2019 pandemic in Hangzhou, China. J Infect [Internet]. 343 2022 Jan;84(1):e9–10. Available from: hcps://doi.org/10.1016/j.jinf.2021.09.007 344 7. Burnec E, Parashar UD, Winn A, Tate JE. Trends in Rotavirus Laboratory DetecMons 345 and Internet Search Volume Before and Aker Rotavirus Vaccine IntroducMon and in 346 the Context of the Coronavirus Disease 2019 Pandemic-United States, 2000-2021. J 347 Infect Dis [Internet]. 2022 Sep 21;226(6):967–74. Available from: 348 hcp://www.ncbi.nlm.nih.gov/pubmed/35184198 349 8. Maldonado -Barrueco A, García-Rodríguez J, Yániz-Ramirez J, Serrano-Vaquero I, 350 Parra-Alonso JC, Vega-Nieto C, et al. Impact of the SARS-CoV-2 Pandemic on the 351 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 24, 2023. ; https://doi.org/10.1101/2023.07.24.23293059doi: medRxiv preprint 18 Prevalence and Incidence of GastrointesMnal Viruses in Children up to Five Years Old: 352 a RetrospecMve Cohort Study. Microbiol Spectr. 2022;10(3). 353 9. Ahn SY, Park JY, Lim IS, Chae SA, Yun SW, Lee NM, et al. Changes in the Occurrence of 354 GastrointesMnal InfecMons aker COVID-19 in Korea. J Korean Med Sci. 2021;36(24):1–355 9. 356 10. Douglas A, Sandmann FG, Allen DJ, Celma CC, Beard S, Larkin L. Impact of COVID-19 357 on naMonal surveillance of norovirus in England and potenMal risk of increased 358 disease acMvity in 2021. J Hosp Infect. 2021;112:124–6. 359 11. KambhampaM AK, Wikswo ME, Barclay L, Vinjé J, Mirza SA. Norovirus Outbreaks 360 Reported Through NoroSTAT — 12 States, August 2012–July 2022. MMWR Recomm 361 Reports. 2022;71(38):1222–4. 362 12. Lu Y, Zhang Z, Xie H, Su W, Wang H, Wang D, et al. The Rise in Norovirus-Related 363 Acute GastroenteriMs During the Fight Against the COVID-19 Pandemic in Southern 364 China. Front Public Heal. 2022;9(January):1–7. 365 13. Zhang W, Wu Y, Wen B, Zhang Y, Wang Y, Yin W, et al. Non-pharmaceuMcal 366 intervenMons for COVID-19 reduced the incidence of infecMous diseases: a controlled 367 interrupted Mme-series study. Infect Dis Poverty [Internet]. 2023;12(1):15. Available 368 from: hcps://doi.org/10.1186/s40249-023-01066-3 369 14. AgoM CN, Curran MD, Murunga N, Ngari M, Muthumbi E, Lambisia AW, et al. 370 Differences in epidemiology of enteropathogens in children pre- and post-rotavirus 371 vaccine introducMon in Kilifi, coastal Kenya. Gut Pathog [Internet]. 2022 Dec 372 1;14(1):32. Available from: 373 hcps://gutpathogens.biomedcentral.com/arMcles/10.1186/s13099-022-00506-z 374 15. Lambisia AW. Comparison of the DiagnosMc Performance of TaqMan Array Cards, 375 Enzyme Immunoassay, Real-Time PCR and Next GeneraMon Sequencing in 376 InvesMgaMon of Five Common Diarrhoea-Associated Enteric Viruses in Kilifi, Kenya 377 [Internet]. JKUAT-COHES; 2021. Available from: 378 hcp://ir.jkuat.ac.ke/handle/123456789/5621 379 16. Lambisia AW, Onchaga S, Murunga N, Lewa CS, Nyanjom SG, AgoM CN. 380 Epidemiological Trends of Five Common Diarrhea-Associated Enteric Viruses Pre- and 381 Post-Rotavirus Vaccine IntroducMon in Coastal Kenya. Pathogens [Internet]. 2020 Aug 382 15;9(8):660. Available from: hcps://www.mdpi.com/2076-0817/9/8/660 383 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 24, 2023. ; https://doi.org/10.1101/2023.07.24.23293059doi: medRxiv preprint 19 17. Nokes DJ, Abwao J, Pamba A, Peenze I, Dewar J, Maghenda JK, et al. Incidence and 384 Clinical CharacterisMcs of Group A Rotavirus InfecMons among Children Admiced to 385 Hospital in Kilifi, Kenya. PLOS Med [Internet]. 2008 Jul 22;5(7):e153. Available from: 386 hcps://doi.org/10.1371/journal.pmed.0050153 387 18. Lambisia AW, M akori TO, Mutunga M, Cheruiyot R, Murunga N, Quick J, et al. 388 Genomic epidemiology of human adenovirus F40 and F41 in coastal Kenya: A 389 retrospecMve hospital-based surveillance study (2013-2022). Virus Evol. 390 2023;9(1):vead023. 391 19. Makori TO, Bargul JL, Lambisia AW, Mwanga MJ, Murunga N, de Laurent ZR, et al. 392 Genomic epidemiology of the rotavirus G2P[4] strains in coastal Kenya pre- and post-393 rotavirus vaccine introducMon, 2012–8. Virus Evol [Internet]. 2023 Jan 394 1;9(1):vead025. Available from: hcps://doi.org/10.1093/ve/vead025 395 20. Vennema H, Kroneman A. Rotavirus A Genotyping Tool [Internet]. [cited 2022 Apr 396 15]. Available from: hcps://www.rivm.nl/mpf/typingtool/rotavirusa/ 397 21. Mathieu E, Ritchie H, Rodés-Guirao L, Appel C, Giamno C, Hasell J, et al. Coronavirus 398 Pandemic (COVID-19). Our World Data. 2020; 399 22. Brand SPC, Ojal J, Aziza R, Were V, Okiro EA, Kombe IK, et al. COVID-19 transmission 400 dynamics underlying epidemic waves in Kenya. Science (80- ) [Internet]. 2021 Nov 401 19;374(6570):989–94. Available from: 402 hcps://www.science.org/doi/10.1126/science.abk0414 403 23. Lewis K. Vesikari Clinical Severity Scoring System Manual. Path [Internet]. 404 2011;(May):1 –50. Available from: 405 hcps://www.path.org/publicaMons/files/VAD_vesikari_scoring_manual.pdf 406 24. Health M of. First case of coronavirus disease confirmed in Kenya. Press Release. 407 2020;1–3. 408 25. Standard. Newborns at risk as shortage of the rotavirus vaccine is felt in various 409 clinics [Internet]. [cited 2023 Jun 2]. Available from: 410 hcps://www.standardmedia.co.ke/health/health-411 science/arMcle/2001448711/newborns-at-risk-as-shortage-of-the-rotavirus-vaccine-412 is-felt-in-various-clinics 413 26. Amit LN, John JL, Mori D, Chin AZ, Mosiun AK, Ahmed K. Increase in rotavirus 414 prevalence with the emergence of genotype G9P[8] in replacement of genotype 415 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 24, 2023. ; https://doi.org/10.1101/2023.07.24.23293059doi: medRxiv preprint 20 G12P[6] in Sabah, Malaysia. Arch Virol [Internet]. 2023;168(6):173. Available from: 416 hcps://doi.org/10.1007/s00705-023-05803-9 417 27. Zhang SX, Zhou YM, Xu W, Tian LG, Chen JX, Chen SH, et al. Impact of co-infecMons 418 with enteric pathogens on children suffering from acute diarrhea in southwest China. 419 Infect Dis Poverty [Internet]. 2016 Dec 27;5(1):64. Available from: 420 hcp://idpjournal.biomedcentral.com/arMcles/10.1186/s40249-016-0157-2 421 422 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 24, 2023. ; https://doi.org/10.1101/2023.07.24.23293059doi: medRxiv preprint