The impact of the COVID-19 pandemic on HIV, malaria, tuberculosis and childhood vaccination services in Guinea: an interrupted time series analysis Authors: Alhassane Diallo, Sidibe Sidikiba, Aboubacar Sidiki Cissé, Mohamed Camara, Mohamed Kaba, Boubacar Djelo Diallo, Miguel Carlos-Bolumbu, Fodé Bangaly Sako, Alexandre Delamou Published: 2025 ABSTRACT AbstractBackgroundEvidence impact of the COVID-19 pandemic on health-related indicator in Sub-Saharan Africa is limited. We aimed to assess the indirect effect of COVID-19 on essential health service and outcomes in Guinea.MethodsIn this interrupted time series analysis, we analysed a nationally surveillance data from January 2018 to December 2022 of nine indicators of HIV, malaria, tuberculosis (TB) and childhood vaccination. We fitted a Poisson segmented regression model accounting for seasonality to estimate the immediate impact of the COVID-19 on these outcomes as well as per-pandemic trend through incidence rate ratio (IRRs) with 95% CIs.ResultsBetween January 2018 and December 2022, the month-to-month (quarter-to-quarter for TB) changes before COVID-19 outbreak increased from 0.4% to 6.4% in all indicators except for TB therapeutic success rate (IRR 0.995, 95% CI 0.987 to 1.004). Overall, there was a decrease in three indicators ranging from 6% for pentavalent vaccine (IRR 0.940, 0.906 to 0.974) to 15% for TB total cases notification (IRR 0.850, 0.785 to 0.920). The pandemic COVID-19 trend significantly downward monthly for four indicators ranging from 0.8% for in-patient malaria cases (IRR 0.992, 0.986 to 0.998) to 8% for PCR test in infants (IRR 0.920, 0.902 to 0.938), and significantly increased monthly by 2.5% for TB-positive microscopy (IRR 1.025, 1.015 to 1.036) and by 0.9% for TB therapeutic success rate (IRR 1.009, 1.001 to 1.017).ConclusionIn Guinea, during the 2020 COVID-19 pandemic, the malaria indicators were generally maintained, while the number of HIV tests in infants, childhood vaccines and TB indicators were heavily impacted. There is an urgent need for more protective and targeted strategies to improve the preparedness of the healthcare service.Keywords: Africa South of the Sahara, Global Health, HIV, Malaria, Tuberculosis WHAT IS ALREADY KNOWN ABOUT THIS TOPIC The COVID-19 pandemic has disrupted health systems globally, particularly in low-income and middle-income countries. The COVID-19 pandemic has disrupted health systems globally, particularly in low-income and middle-income countries. Modelling studies predicted that essential health services such as HIV care, tuberculosis (TB) programmes, malaria control and childhood immunisation could be severely affected by pandemic response measures. Modelling studies predicted that essential health services such as HIV care, tuberculosis (TB) programmes, malaria control and childhood immunisation could be severely affected by pandemic response measures. However, there is limited empirical evidence from West African countries, especially Guinea, using real-world data to assess the magnitude and duration of these disruptions. However, there is limited empirical evidence from West African countries, especially Guinea, using real-world data to assess the magnitude and duration of these disruptions. WHAT THIS STUDY ADDS This study provides nationally representative, longitudinal evidence on how the COVID-19 pandemic impacted nine essential health indicators in Guinea from 2018 to 2022. This study provides nationally representative, longitudinal evidence on how the COVID-19 pandemic impacted nine essential health indicators in Guinea from 2018 to 2022. It reveals a significant immediate decline in TB case notification, infant HIV testing and childhood immunisation, while malaria indicators were relatively stable. It reveals a significant immediate decline in TB case notification, infant HIV testing and childhood immunisation, while malaria indicators were relatively stable. HOW MIGHT THIS STUDY AFFECT RESEARCH, PRACTICE OR POLICY This study highlights the importance of integrating health system resilience into epidemic preparedness plans to maintain essential services during health crises. This study highlights the importance of integrating health system resilience into epidemic preparedness plans to maintain essential services during health crises. The findings can inform targeted policy responses in Guinea and similar settings to prioritise recovery strategies for the most affected health programmes. The findings can inform targeted policy responses in Guinea and similar settings to prioritise recovery strategies for the most affected health programmes. It underscores the need for continuous surveillance and real-time data use to anticipate service disruptions and support evidence-based decision-making in future emergencies. It underscores the need for continuous surveillance and real-time data use to anticipate service disruptions and support evidence-based decision-making in future emergencies. Background The COVID-19 pandemic has significantly disrupted health systems globally, particularly in low and middle-income countries, where fragile healthcare infrastructures and limited resources pose challenges to maintaining essential health services. In sub-Saharan Africa, where the burden of communicable diseases such as HIV, malaria, tuberculosis (TB) and vaccine-preventable diseases remains high, the indirect effects of the pandemic threaten to reverse decades of progress in disease control and prevention efforts.1 2 Emerging evidence indicates that pandemic response measures, including lockdowns, curfews and the reallocation of healthcare resources, have resulted in reduced access to routine healthcare, diagnostic delays and decreased health-seeking behaviours.3 4 Disruptions in immunisation services, HIV testing and treatment, TB diagnosis and malaria case management have been reported across various African countries, with modelling studies predicting increases in morbidity and mortality.5 6 For example, in 65 South African primary care clinics, HIV testing and initiating antiretroviral therapy (ART) were severely impacted during the 2020 COVID-19 lockdown, while ART provision was generally maintained.7 Similarly, in Kenya, the COVID-19 pandemic had various indirect effects on the use of ambulatory health services, including a significant negative impact on outpatient consultations and antenatal care visits nationwide.8 In Togo, HIV, malaria and TB services were generally maintained over time despite the COVID-19 pandemic.9 Additionally, in Ghana, TB case notifications, HIV testing and counselling services were negatively impacted during the COVID-19 pandemic.10 However, real-world data on these disruptions remain scarce, particularly in Guinea, where health systems have been previously strained by major epidemics such as Ebola. To this end, most maternal and child healthcare indicators have been shown to significantly decline during the Ebola virus disease outbreak in 2014.11 Similar pathway was seen for the number of TB cases notified and TB treatment outcome.12 In Guinea, the COVID-19 pandemic is responsible for 38 210 infections and 455 deaths.13 Understanding the magnitude and duration of COVID-19’s impact on essential health services is critical for strengthening epidemic preparedness and designing targeted recovery strategies. In this study, we aimed to evaluate the indirect effects of the COVID-19 pandemic on the delivery of HIV, TB, malaria and childhood immunisation services in Guinea. Methods Study design We did an uncontrolled interrupted time series analysis14 using a nationally surveillance data from January 2018 to December 2022 of nine indicators of HIV, malaria, TB and childhood immunisation. The TB data were collected by quarter (quarter 1 2018 to quarter 4 2022; 20 quarters), while those from the other indicators were monthly recorded. Setting and analysis time point Guinea is located in West Africa, with an estimated population of 13.6 million in 2023, the majority of whom are women; 44% of the population is under 15 years of age, and 37% lives in urban areas.15 In Guinea, the health system divides into eight prefectural health districts, which cover services provided nationwide and contain collected health services data. The National Health Information System (DHIS 2) covered data of 3073 primary health facilities, 26 prefectural hospitals, 8 regional hospitals and 3 national hospitals.16 These data are regularly cleaned and validated by the Ministry of Health division in charge of data monitoring. As the first case of COVID-19 was reported in March 2020, we defined the prepandemic (baseline) period from January 2018 to March 2020 (27 months or 9 quarters for TB), and the pandemic period from April 2020 to December 2022 (33 months or 11 quarters for TB). Data sources and management During the study period, we extracted data using standardised data abstraction forms from the DHIS 2, which contains a module on the collection and analysis of health service indicators after validation by the division in charge of data monitoring. For each health service indicator, we used aggregated data representing the number of events per quarter (for the TB data) and month (for malaria, HIV and immunisation data). The TB indicators were collected by TB-treating health facilities that report to the National Tuberculosis Control Program. To ensure data quality, we performed data collection process for each indicator independently by two trained reviewers. Thereafter, a multidisciplinary team including stakeholders’ programme for TB, malaria, HIV and immunisation, and epidemiologists validated each indicator. Outcomes The main outcomes were indicators related to malaria, HIV, TB and childhood immunisation health services. A total of nine indicators, three for each malaria and TB control, two for childhood immunisation and one for HIV, were considered. These included (1) the number of infants born to HIV-positive mothers who underwent PCR testing at 6 weeks for HIV, (2) the number of children younger than 12 months who had completed their pentavalent vaccination schedule (three doses combined with Haemophilus influenzae type b vaccine), and (3) BCG (anti-TB vaccine) vaccine for childhood immunisation. The indicators considered for malaria were (4) the number of malaria cases confirmed by rapid diagnostic test (RDT) or thick smear, (5) the number of patients treated with artemether and lumefantrine and (6) the number of patients hospitalised for malaria. For TB control services, the indicators were (7) the number of positive microscopies, (8) the number of total TB case notifications and (9) the therapeutic TB success rate, defined as the percentage of reported TB patients who have been successfully treated. Statistical analysis To estimate the immediate and pandemic trend impact of the COVID-19 on these outcomes, we fitted a Poisson segmented regression model with Newey-West SEs17 to accommodate for serial correlation and heteroscedasticity, and adjusted for any potential effect of birth seasonality in outcomes. The Poisson model for the interrupted time series analysis followed the format: where Y is indicator of interest, ß0 estimates the number of individuals using the service at the beginning of the COVID-19 outbreak period; ß1 estimates the average change (pre-trend period) in log(E(Y)) per period over the pre-COVID-19 outbreak pandemic; Time is the number of time points since the beginning of the observation period (up to 20 for TB indicators and 60 for the remaining indicators); ß2 estimates the average change in level of log(E(Y)) that occurred in the period immediately after the COVID-19 outbreak; ß3 estimates the average difference in the slope (trend) during the COVID-19 pandemic compared with the pre-COVID-19 period; Time after COVID-19 is the number of time points since the declaration of the pandemic (0 for the prepandemic period); ß4 to ß7 account for seasonality by including into Poisson model two pairs of sine and cosine terms (Fourier terms). To assess the potential impact of seasonal change in outcome, we did sensitivity analyses by dropping Fourier terms from the Poisson model. Incidence rate ratios (IRRs) with their 95% CIs were used as association measures. IRR<1 indicates a downward (negative) in level or trend, and IRR>1 indicates an upward (positive) in level or trend. To address one of the reviewers’ relevant questions about the decline in the number of children vaccinated with BCG vaccine from August 2022 (figure 1) due to the missing data in several prefectural health services, we used an outliers modelling approach by adding in the Poisson segmented model a dummy variable at affected time point, namely the period from August to November 2022 when the extreme values occurred to account for the potential outliers influence on parameter estimates (as shown in equation 2 from online supplemental appendix). Therefore, the dummy variable was coded as 1 during period with extreme values and 0 otherwise. The coefficient for the outlier indicator quantifies the average deviation of these outliers from expected values, adjusting for time trend and the COVID-19 pandemic effects. The positive value of the coefficient for the outlier indicator means that extreme values are an average higher than expected, and the negative value means that extreme values are an average lower than expected. This approach allows us to adjust for extreme values, model fluctuation without distorting the main parameter estimates (ß2 and ß3), and to avoid the loss of information by excluding these values in the model. Differences were considered statistically significant at p<5%, and all analyses were performed using R software (V.4.4.1). Poisson segmented regression analysis accounting of seasonality using two Fourier pairs of HIV, malaria, tuberculosis and childhood vaccinations indicators before and after the COVID-19 pandemic in Guinea. AL, Arthemeter and Lumefantine; BCG, Bilié Calmette et Guérin vaccin; PENTA, pentavalent vaccine; RDT, rapid diagnostic test; TB, tuberculosis; TS, thics smear. Patient and public involvement Patients or the public were not involved in the design, or conduct, or reporting or dissemination plans of our research. Results Characteristics of health services indicators Between January 2018 and December 2022, 2 206 262 infants received one dose of BCG vaccine and 2 023 490 completed their third dose of pentavalent vaccine. The total number of malaria cases confirmed by RDT was 9 573 282, in whom 9 028 808 cases were treated with artemether and lumefantrine, and 786 706 patients were hospitalised for malaria. A total of 83 811 TB cases were reported, in whom 47 418 were diagnosed with positive microscopies, and the therapeutic TB success ranged from 75% in the first quarter of 2021 to 92% in the third quarter 2022. For the HIV indicator, 6 310 HIV tests were done in infants born to HIV-positive mothers at 6 weeks. Overall, the distribution of service delivery indicators varied between periods before and during the COVID-19 pandemic. Except for the number of children vaccinated with BCG, indicator values were on average higher during the COVID-19 pandemic than before (table 1). Table 1HIV, malaria, tuberculosis and childhood vaccination services before and during the COVID-19 pandemic in Guinea, January 2018 to December 2022CharacteristicsObservation periodPre-COVID-19During COVID-19Overall HIV indicators N=27 months N=33 months N=60 months PCR test in infants Mean (SD)96 (59)112 (34)105 (47) Median (IQR)80 (48–146)113 (86–133)106 (68–136) Range20–22956–19520–229 Malaria indicators N=27 months N=33 months N=60 months Malaria confirmed case (RTD) Mean (SD)128 723 (43562)184 780 (53033)159 555 (56148) Median (IQR)111 937 (91 056–166 780)183 494 (1 31 889–2 28 573)155 733 (1 13 428–1 98 098) Range71 699–216 89594 311–285 54671 699–285 546Patients treated with AL Mean (SD)120 901 (41 162)174 681 (50 901)150 480 (53 664) Median (IQR)104 643 (85 364–157 254)173 100 (125 249–215 377)146 142 (104 244–187 018) Range68 258–204 12089 772–270 18168 258–270 181In-patient malaria cases Mean (SD)12 297 (3434)13 778 (3469)13 112 (3504) Median (IQR)11 443 (9055–15 422)14 143 (11 180–16 288)13 972 (9753–15 961) Range7174–17 6558056–21 5007174–21 500 Childhood indicators N=27 months N=33 months N=60 months BCG vaccination Mean (SD)37 852 (2298)35 887 (8239)36 771 (6334) Median (IQR)37 980 (36 112–39 666)37 839 (35 982–40 124)37 932 (36 035–39 845) Range32 294–42 0113825–43 6293825–43 629Pentavalent vaccination Mean (SD)33 214 (2326)34 143 (1348)33 725 (1894) Median (IQR)32 918 (31 791–34700)34 119 (33 538–34838)33 852 (32 776–34740) Range28 864–40 12731 021–37 65628 864–40 127 TB indicators N=9 months N=11 months N=20 months Positive microscopy Mean (SD)2162 (150)2542 (346)2371 (332) Median (IQR)2189 (2041–2236)2596 (2418–2754)2336 (2116–2602) Range1974–24131819–30151819–3015TB cases notification Mean (SD)3851 (388)4468 (603)4191 (595) Median (IQR)3791 (3456–4060)4572 (4272–4867)4252 (3738–4581) Range3456–44273130–52453130–5245TB therapeutic success rate Mean (SD)86 (4)87 (4)87 (4) Median (IQR)87 (85–88)88 (86–89)87 (86–89) Range79–9175–9275–92*BCG: Bilié Calmette et Guérin vaccin; TS: thics smear, AL: Arthemeter and LumefantineAL, Arthemeter and Lumefantine; BCG, Bilié Calmette et Guérin vaccin; RDT, rapid diagnostic test; TB, tuberculosis. HIV, malaria, tuberculosis and childhood vaccination services before and during the COVID-19 pandemic in Guinea, January 2018 to December 2022 BCG: Bilié Calmette et Guérin vaccin; TS: thics smear, AL: Arthemeter and Lumefantine AL, Arthemeter and Lumefantine; BCG, Bilié Calmette et Guérin vaccin; RDT, rapid diagnostic test; TB, tuberculosis. Immediate impact after the declaration of the pandemic Immediately after the declaration of the COVID-19 pandemic, there was a decrease by 6% for pentavalent vaccine (IRR 0.940, 0.906 to 0.974), 10% for the number of positive microscopies (IRR 0.900, 0.828 to 0.977), and 15% for TB total cases notification (IRR 0.850, 0.785 to 0.920). Conversely, for HIV and malaria indicators and the therapeutic TB success remained constant after the COVID-19 outbreak (table 2 and figure 1). The auto and partial cross-correlation test of indicators shows a significant lag for HIV, malaria, the number of total TB case notifications and the therapeutic TB success rate (onlinesupplemental Figures S1S3). Table 2Poisson segmented regression models accounting of seasonality using two Fourier pairs of the impact of COVID-19 pandemic on HIV, malaria, tuberculosis and childhood vaccination services in GuineaIRR slope change during pandemic (ß3)IRR level change at pandemic (ß2)Pre-pandemic trend (ß1)HIV indicators PCR test in infants0.920 (0.902–0.938)0.808 (0.593–1.100)1.064 (1.044–1.085)Malaria indicators Malaria confirmed case (RTD) 0.993 (0.988–0.998) 1.004 (0.901–1.119) 1.014 (1.011–1.018) Patients treated with AL 0.993 (0.988–0.998) 1.019 (0.905–1.146) 1.014 (1.010–1.018) In-patient malaria cases 0.992 (0.986–0.998) 0.944 (0.846–1.055) 1.008 (1.006–1.011) Childhood indicators BCG vaccination*0.995 (0.988–1.002)0.958 (0.852–1.078) 1.005 (1.000–1.011) Pentavalent vaccination0.997 (0.995–1.000) 0.940 (0.906–0.974) 1.004 (1.002–1.007) TB indicators Positive microscopy 1.025 (1.015–1.036) 0.900 (0.828–0.977) 1.012 (1.006–1.021) TB cases notification1.007 (0.995–1.019) 0.850 (0.785–0.920) 1.028 (1.023–1.032) TB therapeutic success rate 1.009 (1.001–1.017) 0.992 (0.955–1.030)0.995 (0.987–1.004)Boldface value signifies p-value < 0.05 or 95% CI did not included 1.*BCG vaccine was modelled using modelling outlier approach because of extreme value between August and November 2022. The dummy variable for period with extreme values was added in the Poisson regression model to account for the potential outliers influence on parameter estimates. The estimated value of the outlier parameter in the model was 0.434 (IRR=0.435; 95% CI, 0.358 to 0.528), suggesting that extreme values decreased by 57% at affected time point than those expected.IRR, incidence rate ratio; RDT, rapid diagnostic test; TB, tuberculosis. Poisson segmented regression models accounting of seasonality using two Fourier pairs of the impact of COVID-19 pandemic on HIV, malaria, tuberculosis and childhood vaccination services in Guinea Boldface value signifies p-value < 0.05 or 95% CI did not included 1. BCG vaccine was modelled using modelling outlier approach because of extreme value between August and November 2022. The dummy variable for period with extreme values was added in the Poisson regression model to account for the potential outliers influence on parameter estimates. The estimated value of the outlier parameter in the model was 0.434 (IRR=0.435; 95% CI, 0.358 to 0.528), suggesting that extreme values decreased by 57% at affected time point than those expected. IRR, incidence rate ratio; RDT, rapid diagnostic test; TB, tuberculosis. Pandemic trend During study period, the month-to-month (quarter-to-quarter for TB) changes before COVID-19 outbreak increased from 0.4% to 6.4% in all indicators except for TB therapeutic success rate (IRR 0.995, 95% CI 0.987 to 1.004; table 2 and figure 1). The pandemic COVID-19 trend significantly downward monthly for four indicators ranging from 0.8% for in-patient malaria cases (IRR 0.992, 0.986 to 0.998) to 8% for PCR test in infants (IRR 0.920, 0.902 to 0.938) and significantly increased monthly by 2.5% for TB-positive microscopy (IRR 1.025, 1.015 to 1.036) and by 0.9% for TB therapeutic success rate (IRR 1.009, 1.001 to 1.017). Findings from outlier modelling for the number of children vaccinated with BCG vaccine As the number of children vaccinated with BCG vaccine declined dramatically from August to November 2022 (figure 1 and online supplemental figure S4) due to the missing data in several prefectural health services, we performed an outlier modelling approach to prevent bias in the estimation of time trend and COVID-19 interruption effects. We found that immediately after the declaration of the COVID-19 pandemic, the number of children vaccinated with BCG remained constant after the COVID-19 outbreak (IRR 0.958, 0.852 to 1.078; table 2, figure 1 and online supplemental figure S4). During the study period, the month-to-month changes before COVID-19 outbreak increased by 0.5% (IRR 1.005, 95% CI 1.000 to 1.011), and the pandemic COVID-19 trend remained constant (IRR 0.995, 0.988 to 1.002; table 2, figure 1 and online supplemental figure S4). The estimated value of the outlier parameter in the model was statistically significant (IRR=0.435; 95% CI 0.358 to 0.528), indicating that observation after July 2022, those identified as extreme values were, on average 57% lower than those expected at affected time points. Sensitivity analyses In sensitivity analyses, which do not consider the impact of seasonality on outcomes, only the number of children receiving pentavalent vaccine remained significantly down immediately after the COVID-19 epidemic. From pandemic trend analyses (pre-lockdown and pandemic trend), the main results remained similar for HIV, TB and childhood immunisation indicators (online supplemental Table S1). Discussion This study provides robust empirical evidence on the indirect effects of the COVID-19 pandemic on essential health services in Guinea. Using nationally representative surveillance data and applying segmented Poisson regression models, we observed that service disruptions were heterogeneous across disease programmes, with significant declines in TB case notification, infant HIV testing and childhood vaccination coverage, but relative maintenance of malaria indicators. The decline in the total TB cases notification during the pandemic is consistent with findings from other low and middle-income countries. WHO reported a global drop of 18% in TB case notifications in 2020, with Africa contributing significantly to the gap.3 In Guinea, the 15% decrease in TB cases likely reflects disruptions in health service access, reduced community-based screening and patients’ reluctance to seek care due to fear of SARS-CoV-2 exposure trends mirrored in neighbouring countries.18 In particular, the drop in TB notifications immediately after the COVID-19 pandemic was greatest in Togo (27%) and Ghana (32%).9 10 In contrast, TB therapeutic success rate improved slightly during and after the pandemic. This may reflect a survivor or reporting bias, whereby only patients able to access care under difficult conditions completed treatment. It may also result from strengthened follow-up efforts in specific programmes or improved adherence during lockdowns, a finding that requires further qualitative exploration. In this line, the reported TB therapeutic success rate reached 92% in the third quarter 2022. Infant HIV PCR testing declined both immediately after the pandemic onset and in subsequent months. Early infant diagnosis relies on timely health facility visits and laboratory capacity two elements severely affected during health crises. Our findings align with previous studies from South Africa, Kenya, Togo and Ghana, showing sharp declines in early infant diagnosis services and paediatric ART initiation during COVID-19.7,10 Such delays in testing and treatment may have lasting consequences on paediatric HIV morbidity and mortality. There is needed that strategies to increase HIV testing should be implemented to address the future outbreaks, including efforts to improve early infant diagnosis, especially in rural areas where access to the HIV PCR testing and point-of-care diagnosis remains limited. Childhood immunisation, specifically pentavalent vaccine coverage, also declined. Disruptions in routine immunisation programmes during the pandemic have been widely documented, with WHO estimating that over 25 million children missed one or more doses in 2021 alone.19 Additionally, in Mexico, child vaccinations declined by 36%.20 In Guinea, as in many countries, temporary closure of vaccination sites particularly in some rural districts due to the movement restriction and redeployment of health personnel, and lack of transport, likely contributed to reduced uptake. Furthermore, evidence from the 2014–2016 Ebola epidemic outbreak documented reduced care-seeking behaviour due to fear of nosocomial transmission.21 In addition, report during COVID-19 pandemic also reported reluctance to visit facilities and social distress, contributing to reduced vaccine uptake. Interestingly, malaria indicators remained largely stable, reflecting the resilience of malaria control programmes, conversely to the 2014–2016 Ebola epidemic where evidence from the data gathering in health facilities, confirmed through weekly surveillance data, indicate a substantial decline in all-cause outpatient visits, fever cases and number of patients treated with antimalarial drugs at public health facilities.22 This same resilience has been observed in Togo, with a significant decrease in the levels of all malaria indicators immediately after the onset of COVID-19, and a constant trend observed during the pandemic.9 This may be due to several factors. First, the Guinean health system rapidly identified malaria as an essential service at the onset of the COVID-19 pandemic, and national guidance explicitly prioritised continuity of diagnosis, treatment and vector campaigns. Second, the integration of malaria services into community-based care, sustained supply chains for antimalarials or overlapping symptoms of malaria and COVID-19 prompting continued diagnostic activity.23 Third, campaigns such as insecticide-treated net and seasonal malaria chemoprevention were continually adapted to door-to-door delivery, often enhanced by digital microplanning, which helped sustain or even expand coverage in Guinea. Nonetheless, ongoing vigilance is warranted given that indirect impacts on malaria outcomes, including delayed diagnosis leading to more severe malaria cases, missed opportunities for intermittent preventive treatment during pregnancy or reduced coverage of insecticide-treated net and seasonal malaria chemoprevention campaigns may emerge with time, especially in vulnerable subgroups, such as children under 5 years, pregnant women and older people with comorbidities. Although the postpandemic trend would have made it possible to identify the evolution (slight increase or decrease) of indicators after the pandemic, they were not estimated in this work. During the Ebola epidemic, maternal and child services and TB surveillance were negatively impacted.11 12 This raises questions about the lessons learnt from the Ebola epidemic about the resilience of the health system in Guinea and its preparedness for future epidemics. Furthermore, a country like Guinea is not immune to the double occurrence of epidemics and pandemics such as Ebola and COVID-19. Further studies are needed to assess the efficiency of the Guinean healthcare system in the face of such an eventuality. Strengthening community outreach, reinforcing service delivery at the primary care level and investing in health system resilience will be crucial to prevent long-term consequences. Evidence from scoping review about learning from the COVID-19 pandemic for future epidemics and pandemics preparedness and response in Guinea indicates improvement of health system resilience.24 First, the gains of the post 2014–2016 Ebola health system reforms that led to the establishment of a central public health emergency operation department networked with 38 health districts and the training of biosecurity workforce allowed an early response, about 47 days, in face of the SARS-CoV-2 introduction on the national territory. Second, the declaration of the COVID-19 pandemic triggered the undertaking and implementation of several innovative approaches, including a phytomedicine therapeutic trial, manufacturing of hydroalcoholic solutions and traditional facemasks, and COVID-19-patient monitoring and spraying machines. Third, the COVID-19 pandemic response in Guinea acknowledges the rise of new players in the health system compared with Ebola in 2014/2016.24 Furthermore, more concrete examples of targeted recovery strategies were implemented by health authorities to mitigate disruption in HIV, TB and vaccination services. These included mobile HIV testing campaigns, community-based ART delivery, decentralised drug distribution points and catch-up immunisation campaigns for children who missed routine vaccination. Despite the encouraging discussed, some challenges remain. These include human resources by training, protection and retention of frontline health workers, especially the community health workers; the supply chains by strengthening logistics and stock management systems; the primary care infrastructure by ensuring that rural health posts are adequately equipped by oxygenators or respirators, and staffed; the health information system by improving data collection and real-time monitoring to detect service disruptions early; and the community engagement by building trust and reinforcing communication to sustain demand for services during crises. Moreover, evidence indicates that services of COVID-19 disruptions were not evenly distributed.24 Rural communities, disability populations and socioeconomically disadvantaged groups were often disproportionately affected due to limited access to health facilities, transport restrictions or digital divide. Targeted recovery efforts will therefore need to explicitly address these inequities by bringing services closer to underserved communities, prioritising outreach in high-burden or hard-to-reach areas, and tailoring interventions to vulnerable groups. Strengths and limitations Our study is based on routine surveillance data and may be subjected to under-reporting or data quality issues, although independent data extraction has mitigated this bias. Moreover, the number of children vaccinated with BCG vaccine after July 2022 decrease dramatically due to the missing data from several prefectural health facilities. As indicated by simulation studies,25 an analysis of interrupted time series based on aggregated-level data with missing at random process systematically biases the key parameter estimation (level and slope changes after the interruption). Simulation studies suggest performing multiple imputations using individual-level data and to estimate the impact of the interruption using mixed random-effects models.25 We accounted for these extreme observations using an outlier modelling approach, which has been recommended as a best practice method in segmented regressions Interrupted time-series (ITS) analysis.14 This approach avoids bias introduced by data deletion or imputation and allows the intervention (the COVID-19 pandemic in our context) effect to be estimated without undue influence from anomalous values. The coefficient associated with the outlier dummy variable from model has no substantive interpretation; instead, it isolates abnormal fluctuations arising from surveillance artefacts, reporting issues or unusual epidemiological events. We found that extreme values were an average 57% lower than those expected at affected time point. Furthermore, the likelihood test ratio comparing models with and without incorporation of the dummy outlier variable into the model was statistically significant (p<0.0001) in favour of the complete model, suggesting a better consideration of the impact of missingness process in our conclusion. We also cannot account for all contextual factors (eg, health worker strikes, funding variations) that may have influenced trends. Additionally, the vaccination coverage is generally more meaningful than the number of children vaccinated. However, given the unreliability of the number of target children for vaccination each month, we used the most reliable and well-documented immunisation indicators such as the number of children vaccinated. Conversely, the long prepandemic time frame and seasonally adjusted models enhance the robustness of our findings. Conclusion This study demonstrates that the COVID-19 pandemic had substantial and variable impacts on essential health services in Guinea. While malaria services were relatively preserved, total TB case detection, infant HIV diagnosis and vaccination coverage were significantly disrupted. These findings call for renewed investment in health system resilience, continuity of care during health emergencies and targeted recovery interventions for the most affected services. Supplementary material Acknowledgements We would like to thank the team at the Guinean Ministry of Health and the managers of the various national tuberculosis, malaria, HIV and immunisation programmes for facilitating access to the data and their collaboration. Footnotes Funding: The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors. Provenance and peer review: Not commissioned; externally peer-reviewed. Handling editor: Gang Qin Patient consent for publication: Not applicable. Ethics approval: Not applicable. Patient and public involvement: Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research. Data availability statement The data are available upon request (djuhany@gmail.com). References Associated Data Supplementary Materials Data Availability Statement The data are available upon request (djuhany@gmail.com).