Absorbing the shock: the role of lower-level health facilities in immunization resilience in Sidama, Ethiopia during COVID-19 pandemic

Abstract

Background

Although the COVID-19 pandemic disrupted routine immunization services globally, evidence from Ethiopia suggests limited overall impact. However, little is known about how immunization uptake varied across different facility levels and phases of the pandemic. This study examined these patterns in public hospitals, health centers, and health posts.

Methods and materials

We conducted a longitudinal comparative study of routine immunization service uptake across 18 hospitals, 132 health centers, and 532 health posts in Sidama Regional State, Ethiopia. To assess disruption and recovery patterns, we compared immunization uptake among infants under one year of age during the first 12 months of the pandemic (March 2020-February 2021) with the corresponding pre-pandemic period (March 2019-February 2020). We extracted data from the District Health Information System (DHIS2) and calculated percent changes with 95% confidence intervals. We compared mean differences in uptake using a paired t-test.

Results

Overall, immunization uptake increased by 9.72% (95% CI: 9.66%-9.79%) (P = 0.006) during the first six months of the pandemic (March-August 2020), compared to the same pre-pandemic period. While hospitals experienced a decline of 5.2% (P = 0.449), health centers and health posts experienced significant increases of 7.96% (P = 0.015) and 11.5% (P = 0.002), respectively. Compared to the six months immediately before the pandemic (September 2019 - February 2020), overall uptake increased by 3.90% (P = 0.144), with a 14.01% decline in hospitals (P = 0.151), and increases of 4.05% (P = 0.275) and 5.45% (P = 0.032) in health centers and health posts, respectively. During the post-lockdown period (September 2020 - February 2021), immunization uptake increased by 5.11% (P = 0.007) compared to the same pre-pandemic period, including rises of 18.7% in hospitals (P = 0.208), 2.5% in health centers (P = 0.031) and 4.6% in health posts (P = 0.004). Compared to the first half of the pandemic, hospital-level immunization uptake increased significantly by 38.15% (p < 0.001). However, overall uptake increased only by 1.16% (p = 0.461), while there were non-significant declines of -0.76% (p = 0.627) in health posts and − 1.45% (P = 0.790) in health centers. Over the full 12-month pandemic period, total uptake increased significantly by 7.35% (p < 0.001) compared to the previous year. Health posts showed the largest gain (7.99%; p < 0.001), followed by health centers (5.2%; p = 0.003); and non-significant rise of hospitals (5.3%; p = 0.376). Uptake of all vaccine types increased annually, except for BCG, which showed a non-significant decline (p = 0.889).

Conclusions

Immunization services in the region remained resilient during the pandemic across all vaccine antigens, primarily due to the capacity of health posts and health centers to absorb increased demand following disruptions in hospitals. These findings underscore the importance of decentralized health systems in maintaining immunization services during pandemics. Strengthening linkages across facility levels and minimizing early hospital disruptions are essential to ensuring service continuity.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12913-026-15386-1.

Introduction

Vaccine-preventable diseases remain a major cause of global morbidity and mortality [1, 2]. In 2018 alone, vaccine preventable diseases accounted for 700,000 deaths among children under five were reported, most of which occurred in low- and middle-income countries (LMICs) [3]. Childhood immunization is one of the most effective interventions preventing over 20 diseases, averting four million deaths and protecting 750,000 children from disability each year [4, 5]. Immunization contributed to the Millennium Development Goal (MDG) target of reducing child mortality by half between 1990 (90 deaths per 1,000 live births) and 2015 (43 deaths per 1,000 live births) [4–6]. It remains central to achieving the Sustainable Development Goal 3 (SDG3) targets and contributes to 14 of the other 16 SDGs [6, 7]. The Immunization Agenda 2030 (IA2030) aims to enhance global immunization by promoting equitable access to vaccines and mitigating the setbacks caused by the COVID-19 pandemic [8].

Global vaccination coverage had been on an increasing trend until the COVID-19 pandemic [9]. Coverage improved rapidly until 2010 [9, 10] and continued to rise gradually until 2019 [9]. In 2019, global coverage for the third dose of the diphtheria, tetanus, and pertussis vaccine (DTP3) and the first dose (DTP1) was 85% and 90%, respectively [8, 9, 11]. In Africa, DTP3 and DTP1 coverage was 74% and 81%, respectively, in the same year [9, 11, 12], whereas in Ethiopia it was 61% and 76%, respectively [13]. However, this promising global progress was severely disrupted by the COVID-19 pandemic, marking the most significant disruption to immunization efforts in over three decades [14–16], with the greatest impact observed in Africa and South-East Asia [16], despite by governments’ efforts to protect immunization services. However, the effectiveness of these measures varied depending on factors such as caseloads, the pre-existing status of immunization services, and other contextual conditions.

Ethiopia reported its first case of COVID-19 case on 13 March 2020, having begun to implement preparedness measures as early as January. The initial cases were mostly among quarantined individuals, but community transmission caused infections to accelerate from June to August 2019 [17–19]. A national lockdown and state of emergency were enforced in April 2020, including school closures, travel restrictions, and public health measures. These were lifted in September [17–19], after which cases peaked between September and November, marking the first wave. Daily cases then plateaued, and a second wave began in late February 2021 [20]. The outbreak pattern in Sidama region, the geographical focus of this study, reflected the national trend [20]. Ethiopia implemented several national interventions to protect and sustain immunization services during the pandemic [21]. In April 2020, the National Comprehensive Handbook of the Management of COVID-19 was published, outlining strategies to maintain essential services such as immunization [21]. These strategies included optimizing service delivery settings, streamlining patient flow, reallocating health workforce capacity, ensuring vaccine supply and availability, and continuously monitoring service delivery. Priority was given to maintaining immunization services at health centers and health posts, while hospitals primarily focused on responding to the pandemic [21].

The extent to which immunization services were disrupted and subsequently recovered varied across and within countries [22]. In Ethiopia, available evidence suggests minimal disruption to routine immunization services. However, these findings are limited in that most studies focused on the national [23, 24] or regional [24, 25] level, were restricted to the first few months of the pandemic [26–29], were limited to urban areas [28] and health centers [30] only, and did not report separately for different levels of health facilities [31]. Even one to two years after the WHO declared that the public health threat posed by the pandemic had ended, there was still limited information on the disruption to routine immunization services across different levels of public health facilities, and the role of these facilities played in national-level health system resilience. Understanding these disruption and recovery patterns is critical for future outbreak preparedness and response. Lessons learned during the pandemic can help to mitigate service disruption during future outbreaks and prevent long-term consequences [32]. This study examines disruptions to immunization service uptake and recovery patterns across hospitals, health centers, and health posts in Sidama Regional State, Ethiopia, during the first year of the pandemic, when the disruptions were at their most severe.

Materials and methods

Study setting

The study was conducted among all public facilities (18 hospitals, 132 health centers, and 532 health posts) found in Sidama Regional State, one of Ethiopia’s twelve administrative regions [33]. During the study period, the region comprises 32 woredas (districts) and four town administrations, with an estimated population of more than 4 million between 2020 and 2021 [33]. The region’s healthcare system mirrors the national three-tier system, ranging from primary healthcare units (health posts, health centers, and primary hospitals), which serve an estimated 5,000, 25,000, and 100,000 people respectively, to secondary healthcare (general or zonal hospitals), which serve 1–1.5 million people, and tertiary care (referral and specialized hospitals), which serve 3.5–5 million people [34].

Routine child immunization services in Sidama follow Ethiopia’s national immunization schedule, which is part of the Expanded Program on Immunization (EPI). Supported by the government in collaboration with international organizations such as the World Health Organization (WHO), the United Nations Children’s Fund (UNICEF), and Gavi, the Vaccine Alliance (Gavi), the EPI ensures a steady supply of vaccines [13]. During the study period, the Ethiopian Ministry of Health’s routine immunization program included 12 antigens, administered according to the following schedule stated in Table 1 below. Routine immunization services are predominantly delivered through fixed health facilities, accounting for around 80–85% of vaccine administration. The remaining 15–20% is provided via outreach and mobile strategies targeting remote and underserved communities [35, 36].

Table 1.

Routine childhood immunization schedule and vaccine components, during the study period in Sidama region, Ethiopia

Age	Vaccine	Antigens / Target Diseases
At Birth	BCG	Tuberculosis
	OPV (Birth Dose)	Poliovirus (oral, live attenuated)
6 Weeks	OPV1	Poliovirus
	Penta1	Diphtheria, Tetanus, Pertussis, Hepatitis B, Haemophilus influenzae type b
	PCV1	Streptococcus pneumoniae (pneumococcus)
	ROTA1	Rotavirus
10 Weeks	OPV2	Poliovirus
	Penta2	Diphtheria, Tetanus, Pertussis, Hepatitis B, H. influenzae type b
	PCV2	Pneumococcus
	ROTA2	Rotavirus
14 Weeks	OPV3	Poliovirus
	Penta3	Diphtheria, Tetanus, Pertussis, Hepatitis B, H. influenzae type b
	PCV3	Pneumococcus
	IPV	Poliovirus (inactivated)
9 Months	MCV1	Measles virus

Study design and study period

We used a longitudinal comparative design to evaluate changes in the uptake of immunization services during and before the COVID-19 pandemic in all public facilities (health posts, health centers, and hospitals). We collected monthly data over a two-year period (March 2019 to February 2021) on the number of vaccine doses administered across vaccine antigens. This included one year of data before the pandemic (March 2019 to February 2020) and one year of data during the pandemic (March 2020 to February 2021) (Fig. 1).

Fig. 1.

Comparative timeline of routine immunization uptake: Pre-Pandemic vs. During the Pandemic Periods. T1: Immunization uptake during the first six months of the pandemic (March 2020 to August 2020). T2: Immunization uptake during the subsequent six months (September 2020 to February 2021). T3: Total immunization uptake during the pandemic period (March 2020 to February 2021). Baseline 1: Immunization uptake in the six months preceding the pandemic (March 2019 to August 2019). Baseline 2: Immunization uptake in the six months immediately before the pandemic (September 2019 to February 2020)

Study population and outcomes of the study

The study population consisted of infants under 1 year of age who received routine immunization services at public health facilities across the Sidama Regional State, Ethiopia. The primary outcome was the change in routine immunization uptake patterns among infants under one year of age. The magnitude of this change was measured by the percentage change in uptake during the pandemic compared to the pre-pandemic period. The focus was on the national lockdown and the period after it was lifted. The significance of this change was assessed by the mean differences in uptake (monthly, quarterly, and biannually) between the pandemic and pre-pandemic periods. Secondary outcomes included: (1) variation in immunization uptake by type of public health facility and vaccine type; and (2) exploring whether disruption and recovery patterns were associated with the interventions implemented during the pandemic.

Source of data, quality assurance and data validation

The data were extracted from the District Health Information System, version 2 (DHIS2) database. Since 2016, the Ethiopian Ministry of Health has significantly improved the quality and use of routine health data, including immunization data, by rolling out DHIS2 nationwide [37]. This replaced fragmented paper systems with a unified digital platform. DHIS2 data quality, completeness, accuracy, and timeliness, have been improved by implementing various interventions, including the Information Revolution Roadmap, integration with the electronic Community Health Information System (eCHIS), routine data quality audits, capacity building for health staff, and the establishment of Data Use Rooms for regular data review and decision-making [38, 39]. These efforts have led to notable improvements, with data reporting completeness for key indicators such as immunization and maternal health often exceeding 90% [37], although challenges persist in remote areas.

One member of the study team, who is a Health Management Information System (HMIS) expert and co-author, was granted access to the DHIS2 database. Monthly immunization uptake data were retrieved from the regional DHIS2 system using Microsoft Excel. Vaccine uptake figures for each health facility were extracted and aggregated by facility type and vaccine type. For example, the total number of BCG doses administered across 18 hospitals, 132 health centers, and 532 health posts over the 24-month study period was calculated to determine monthly uptake. These monthly aggregates were then summarized into quarterly, biannual, and annual figures in order to assess seasonal variations and examine disruption and recovery patterns. To ensure data confidentiality, only de-identified and aggregated facility-level data were used. No individual patient records or personally identifiable information were accessed. The extracted data were securely stored on password-protected computers, and the findings were reported exclusively in aggregate form. Routine data quality assurance processes, including validation checks for completeness and consistency, had already been implemented by regional health information experts as part of the standard DHIS2 monitoring system. Therefore, the study team did not undertake additional data validation. Missing values were treated as zero uptake for the respective facility and vaccine, based on the assumption that non-reporting reflects no service delivery. This approach aligns with standard DHIS2 data use practices and allows for consistent, conservative estimation of service utilization across the 24-month study period. The eligible population size was not incorporated into our analysis, as the study’s focus was on uptake trends rather than differences in coverage across facilities.

Study variables

The immunization uptake variables were defined as follows:

1. BCG uptake: The number of children under one year of age who received the Bacille Calmette–Guérin (BCG) vaccine.

2. Pentavalent (Penta 1 and 3): The number of children under one year of age who received the combined DPT-Hib-HepB vaccine, administered at 6 and 14 weeks of age.

3. PCV (PCV1 &3): The number of children under one year of age who received the pneumococcal conjugate vaccine, administered at six weeks (PCV1) and 14 weeks (PCV3) of age.

4. Polio (OPV1 & 3): The number of children under one year of age who received the oral poliovirus vaccine at six weeks (OPV1) and 14 weeks (OPV3) of age.

5. Rotavirus (Rota1 &2): The number of children under one year of age who received the rotavirus vaccine at six weeks (Rota1) and 10 weeks (Rota2).

6. IPV uptake: The number of children under one year of age who received one dose of the inactivated poliovirus vaccine, administered at 14 weeks.

7. Measles uptake: The number of children under one year of age who received the first dose of the measles vaccine, administered at nine months.

8. Total or overall uptake: The total number of doses of any of the above vaccines administered per month.

9. Complete vaccination: The number of children under one year of age who received all the recommended doses of vaccine according to the national immunization schedule before their first birthday.

Statistical analysis

The primary outcome of this study was the change in routine immunization uptake among infants under one year of age during the COVID-19 pandemic. We employed two analytical approaches to assess these changes. First, we calculated the magnitude of disruption to uptake and the recovery pattern by the percent change in uptake during pandemic periods (overall, lockdown, and post-lockdown phases) relative to pre-pandemic levels using the formula: [(Pandemic Uptake - Pre-Pandemic Uptake)/Pre-Pandemic Uptake] × 100, with 95% confidence intervals derived through bootstrapping. Secondly, we computed mean differences in monthly, quarterly, and biannual uptake between pandemic and pre-pandemic periods. We employed paired t-tests for normally distributed data and Wilcoxon signed-rank tests for non-parametric distributions in order to assess the significance of the change. Our analyses were stratified across multiple dimensions: by facility level (health posts, health centers, hospitals, and aggregated totals), by vaccine antigen types, and by distinct pandemic phases (Fig. 1). We conducted three temporal comparisons: (1) the initial six-month lockdown period (March-August 2020, T1) versus both the corresponding pre-pandemic period (March-August 2019, Baseline 1) and the six months immediately before lockdown (September 2019-February 2020, Baseline 2); (2) the post-lockdown period (September 2020-February 2021, T2) versus Baseline 2 and T1; and (3) the full pandemic year (March 2020-February 2021, T3) versus the combined pre-pandemic baselines. We assessed the resilience of immunization service uptake during the total pandemic period (compared to pre-pandemic levels) was assessed using Pearson’s or Spearman’s rank correlation, along with their associated p-values to determine statistical significance. A correlation coefficient (r or ρ) close to + 1 indicates strong resilience (minimal disruption), while a value near zero or negative suggests significant disruption in immunization service uptake. A p-value of less than 0.05 was considered statistically significant for all analysis. We visualized the results using comprehensive tables and figures, conducting all analyses with SAS software (version 9.4).

Ethical considerations

We used routine data on immunization uptake before and during the COVID-19 pandemic. The data were sourced from routine vaccine administration reports submitted to DHIS2 and did not include any identifiable patient information, meaning informed consent was not required. We have received ethical approval and a waiver of consent from the Institutional Review Board (IRB) of Hawassa University College of Medicine and Health Sciences (Ref No. IRB/076/13–5 march 2021).

Results

Immunization service uptake pattern during the first half of the pandemic period (National emergency lockdown)

During the first six months of the pandemic (T1: March-August 2020), total immunization uptake across all facilities increased by 9.72% (95% CI: 9.66%-9.79%) compared to the same pre-pandemic period (Baseline 1: March-August 2019), with a mean monthly increase of 14,546.8 doses (p = 0.006). Health posts demonstrated the most substantial increment, with an increase of 11.5% (95% CI: 11.4%-11.6%), equivalent to an additional 12,432.8 doses per month (p = 0.002), followed by health centers with a 7.96% increase (95% CI: 7.8%-8.0%; +2,596.8 monthly doses, p = 0.015). By contrast, hospitals experienced a 5.2% decline (95% CI: -5.0% to -5.4%; 482.8 fewer monthly doses), though this reduction was not statistically significant (p = 0.449) (Fig. 2, T1A; Supplementary Tables 1 & 3).

Fig. 2.

Percentage difference in uptake of routine immunization service six months before the COVID-19 pandemic versus the first 6 months of the pandemic. * T1_A % difference represents the change in immunization uptake during the first six months of the pandemic (T1; March 2020 to August 2020) compared with the same pre-pandemic period (Baseline 1: March 2019 to August 2019). *T2% difference represents the change in immunization uptake during the second half of the pandemic (T2: September 2020 to February 2021), compared with the same pre-pandemic period (Baseline 2: September 2019 to February 2020)

Vaccine-specific analysis revealed that health posts achieved statistically significant increases of over 10% for all vaccines (p < 0.05), while health centers showed significant growth ranging from 7.3% (rotavirus) to 10.04% (BCG) (p < 0.05). Hospitals recorded declines of over 3% for most vaccines, though none of these were statistically significant (p > 0.05 for each vaccine types). The most pronounced reduction was observed for BCG (-12.9%). Compared to the immediate pre-pandemic period (Baseline 2: September 2019-February 2020), total uptake during T1 increased by 3.9% increase (95% CI: 3.86%-3.94%; +6,063.2 monthly doses, p = 0.144). Health posts showed significant increase (+ 5.45%, 95% CI: 5.39%-5.50%; +6,100.7 monthly doses, p = 0.032), whereas health centers exhibited a non-significant increase of 4.05% (95% CI: 3.96%-4.13%, p = 0.275). Hospitals experienced a decline of 14.01% decline (95% CI: -13.73% to -14.28%), with most vaccines decreasing by over 10% (BCG: -17.7%). However, though these reductions remained non-significant (all p > 0.05) (Fig. 3, T1B; Supplementary Tables 1 & 3).

Fig. 3.

Percentage difference in routine immunization service uptake 12 months before and during the COVID-19 pandemic (left) and six months before and during the pandemic (right). * T3% difference shows how immunization uptake changes over a year during the COVID-19 pandemic compared to the year before it started. *T1_B % difference shows the change in immunization uptake during the first half of the pandemic (T1: March-August 2020) compared to the preceding six months before the pandemic (Baseline 2: September 2019 to February 2020)

Immunization service uptake patterns during the second phase of the pandemic (after the national emergency was lifted)

Following the lifting of lockdown measures (T2: September 2020-February 2021), immunization uptake increased across all facilities compared to the immediate pre-pandemic period (Baseline 2: September 2019-February 2020). Overall, there was an increase of 5.11% (95% CI: 5.07–5.15), representing an additional 8,236.2 monthly doses (p = 0.007). This sustained increase was driven by statistically significant increases at both health posts (4.60%, 95% CI: 4.6–4.7; +5,410.0 monthly doses, p = 0.004) and health centers (2.50%, 95% CI: 2.4–2.6; +888.7 monthly doses, p = 0.031). Although hospitals demonstrated an increase of 18.70% (95% CI: 18.4–19.1; +1,937.5 monthly doses), this change was not reach statistically significant (p = 0.208). A vaccine-specific analysis of the hospitals revealed substantial improvements of over 20% for most antigens, with more modest gains for BCG (10%) and measles (4.9%) (Fig. 2, T2; and supplementary Tables 2 & 3). Compared to the first half of the pandemic period (T1), the post-lockdown period showed minimal overall growth (+ 1.16%, p = 0.461), but this masked divergent trends at facility level: hospitals experienced a dramatic 38.15% surge (p < 0.001), while both health posts (-0.76%, p = 0.627) and health centers (-1.45%, p = 0.79) showed non-significant declines.

Overall immunization service uptake pattern during the first 12 months of the pandemic

During the 12-month pandemic period, total immunization uptake increased by 7.35% (95% CI: 7.32–7.39) across all facility levels, with a mean monthly increase of 11,391.5 doses (P < 0.001) compared to the pre-pandemic period. The largest increase was seen at health posts, at 7.99% (95% CI: 7.94–8.03), or 12,401.4 doses per month (P < 0.001). This was followed by, health centers, at 5.20% (95% CI: 5.13–5.27), or 2,737.1 doses per month (p = 0.003). There was a non-significant change at hospitals, at 5.30% (95% CI: 5.20–5.50), or 2,461.5 doses per month (P = 0.376) (Fig. 3, T3; Supplementary Tables s2 & 4).

Comparative analysis of monthly immunization uptake before and during the COVID-19 pandemic revealed significant variations at different levels of health facilities. Health posts demonstrated the most robust performance, with statistically significant increases (p < 0.001) for all vaccine types. The most substantial absolute increase was observed for BCG vaccinations (1,138.7 doses, p < 0.001), followed by the pentavalent and rotavirus vaccines. Complete vaccination uptake increased by 704.3 doses (p = 0.001). Health centers also showed significant increases across all measured antigens (p < 0.05), though with smaller absolute increases compared to health posts. The largest mean difference was exhibited by measles vaccination (209.9 doses, p = 0.001), followed by BCG (223.3 doses, p = 0.009) and pentavalent vaccines. Complete vaccination uptake also increased by 199.4 doses (p = 0.001). Furthermore, a consistent and positive correlation was observed between the two periods at the health center and health posts. In contrast, hospitals presented a markedly different pattern, with no statistically significant changes observed for any individual vaccine (all p > 0.05). BCG vaccinations showed a non-significant decline of 14.8 doses (p = 0.889), while measles vaccinations remained virtually unchanged at 4.3 doses (p = 0.893). The correlation pattern showed weak correlation coefficients (ranging from − 0.26 to 0.52) across all antigens at hospitals BCG antigen exhibited an inconsistent pattern across all facilities, demonstrating a very low to negative correlation compared to other vaccine types (Supplementary Table 4).

Quarterly immunization uptake patterns during the pandemic compared to the pre-pandemic period

As illustrated in Figs. 4 and 5 and Supplementary Table 5, health posts experienced a statistically significant increase in the mean uptake for most vaccines, while health centers showed a non-significant increase during the first quarter of the pandemic (March–May 2020). Hospitals recorded a decrease in all vaccine types, with the sharpest decline in BCG uptake. However, this was not statistically significant. In the second quarter (June–August 2020), overall vaccine uptake increased across health posts (p = 0.043), health centers (p = 0.067), and all facilities combined (p = 0.056). Meanwhile, hospitals continued to experience a decline in most vaccines and total uptake (p = 0.776). Following the lifting of the lockdown in the third quarter (September–November 2020), health centers experienced a slight overall increase in uptake, although there were minor decreases in uptake of early infant vaccines (Penta1, PCV1, OPV1, and Rota1). Health posts reported increased uptake all vaccines (p = 0.12), with a statistically significant rise in BCG uptake (p = 0.008). Hospitals recorded non-significant increases in uptake of all vaccine types, except for BCG, which declined (p = 0.784). Across all facilities, the total immunization uptake increased, though not significantly (p = 0.204). In the final quarter (December 2020–February 2021), overall vaccine uptake improved across all facility types. Statistically significant increases were observed at the regional level (p = 0.015), with near-significant gains were seen at health posts (p = 0.053) and health centers (p = 0.123). However, hospitals showed a non-significant improvement (p = 0.257).

Fig. 4.

Monthly percentage difference in total immunization uptake during the pandemic (March 2020 - February 2021) compared with the pre-pandemic period (March 2019 – February 2020)

Fig. 5.

Quarterly percentage difference in total immunization uptake and number of complete vaccinations during the pandemic (March 2020–February 2021) compared to the pre-pandemic period (March 2019–February 2020). Q1 (early Pandemic): March–May 2020 vs. Mar–May 2019, Q2 (mid-Pandemic): Jun–Aug 2020 vs. Jun–Aug 2019, Q3 (mid Pandemic): Sep–Nov 2020 vs. Sep–Nov 2019, Q4 (late -Pandemic): Dec 2020–Feb 2021 vs. Dec 2019–Feb 2020

Discussion

Despite the widespread disruption anticipated during public health emergencies, our longitudinal analysis reveals a resilient pattern of routine immunization uptake in the Sidama region during the first year of the COVID-19 pandemic. We found a 7.35% increase in overall immunization uptake compared to the pre-pandemic period. However, this overall resilience masked notable disparities across different levels of the health system. While hospitals experienced a significant initial decline of 5.2% in the first six months, primary healthcare facilities effectively compensated for this disruption; health posts and health centers recorded substantial increases of 11.5% and 7.96%, respectively, during the same period. A strong recovery in hospital services in the latter half of the pandemic, coupled with the sustained performance at lower levels, ultimately ensured the continuity of immunization services in the region.

The significant redistribution of immunization uptake during the initial lockdown, with a decline in hospitals and an increase in primary care facilities, and the subsequent post-lockdown reversal reflects the public’s swift adaptation to mobility restrictions. This shift demonstrates the system’s inherent resilience, with lower-level facilities effectively absorbing the demand for hospital services. This is consistent with global reports of healthcare avoidance at tertiary centers during health emergencies. Consistent with our findings, health posts and health centers in Mozambique and Nepal maintained or increased routine vaccination rates by 5–15% during lockdowns, while hospitals experienced an 8–20% decline [40]. A shift in service uptake levels was also observed in Kenya and Lebanon during the COVID-19 pandemic. In Kenya, there was a notable transition from public to private healthcare facilities for immunization services [41, 42]. In Lebanon, there was a shift in service utilization from public to private facilities [43]. Importantly, these inverse trends between hospital and primary care uptake were consistent for all vaccine types, indicating a system-wide shift in demand rather than any vaccine-specific effect.

Understanding the factors that drive changes in vaccine utilization at a facility-level is crucial for strengthening immunization systems and preparing for future outbreaks. The differences observed across facility levels were likely driven by shifts in care-seeking behaviour than actual service disruptions. Early in the pandemic, hospitals became central to COVID-19 screening, diagnosis, and treatment [20], generating widespread fear and discouraging families from seeking immunization services there. Such fear-driven declines in care use, especially early in the pandemic in LMICs are well documented [41]. Movement restrictions, transport barriers, and infection concerns further contributed reductions in hospital-based immunization [42]. Urban areas, where hospitals are concentrated, experienced greater disruptions due to stricter lockdowns [23]. Evidence from Ethiopia’s Oromia region also shows significant hospital-level vaccine supply disruptions [43], likely influenced by health workers’ infection concerns.

Unlike in many other countries, where declines in immunization were primarily linked to vaccine stockouts or staff shortages [44], these issues were less prominent in our study. Lower-level healthcare facilities effectively absorbed the increased demand, indicating that the vaccine supply and distribution remained functional. Hospital disruptions occurred early in the pandemic, before major case surges, and services recovered in the second half of the year despite rising infections. This pattern further supports that shifts in service utilization were driven by pandemic-related behavior changes rather than systemic failures in the healthcare system.

Despite initial disruptions to hospital services, immunization services in the region remained largely resilient, consistent with studies in Ethiopia and other ten countries showing limited overall impact [23–25]. Early declines were followed by rapid recovery, mirroring trends observed in other sub-Saharan African countries [45–47]. Lower-level facilities mitigated the impact of hospital disruptions, with increased vaccine administration across all facility types as seen in Kilifi, Mombasa, and Nakuru counties in Kenya [48, 49]. A similar shift from hospitals to primary care settings was observed in high-income settings. In the United Kingdom, general practitioner practices maintained 93% of pre-pandemic childhood vaccination coverage versus a 42% reduction in hospitals [50];in Ontario, Canada, community health nurses sustained 89% coverage (+ 4% from baseline) compared to a 28% decline in paediatric hospitals [51]; in Australia, Aboriginal Community Controlled Health Organizations achieved 15% higher immunization rates, while urban hospitals experienced a 20% decline [52].

The study findings highlight important lessons for improving outbreak preparedness and response. A key insight is the demonstrated resilience of decentralized primary healthcare, where health posts and health centers effectively maintained and increased immunization uptake during hospital disruptions. This underscores the importance of formally establishing and strengthening primary health care facilities as the cornerstone of emergency response plans through sustained investment in their infrastructure, supply chains, and workforce.

Furthermore, disruption may have been primarily driven by demand-side factors, such as public fear of infection and mobility restrictions, rather than supply-side issues like vaccine shortages and human resource shortages associated with the pandemic prevention. This indicates that preparedness must extend beyond logistics to include proactive, dynamic risk communication and public messaging to reassure communities and direct them towards safe services. The early onset of hospital disruptions, likely triggered by initial response measures, emphasizes that the need for future strategies to be agile enough to cope with pandemics with long lead times and rapidly emerging local outbreaks. Ultimately, building a resilient, multi-level immunization system will not only mitigate the impact of health emergencies, but also contribute to achieving SDGs, including reducing under-five mortality (SDG 3.2) and advancing universal health coverage (SDG 3.8).

Our study has several strengths. First, we assessed the disruption and recovery patterns at all levels of public facilities, providing a better understanding of the role of facility levels in immunization resilience. Secondly, unlike previous studies [26–29] which assessed disruptions in the first six or nine months of the pandemic period, we included data from the first 12 months of the pandemic, helping us to better understand the patterns. Third, we explored the potential impact of the interventions implemented in response to the pandemic on the resilience of immunization services in the event of future similar outbreaks. Finally, our study documented consistency in facility-switching behaviour for the majority of vaccines, providing novel evidence for whole-program adaptation evidence.

Despite its strengths, this study has several limitations. Firstly, our analysis relied on secondary, aggregated monthly data, which may be subject to errors or inconsistencies. Secondly, variations in recording practices across different facilities could affect the accuracy of our findings. Thirdly, our data did not account for individual-level factors that might have influenced vaccine uptake during the pandemic. Fourthly, the study focused solely on public facilities, excluding the limited role of private facilities to maintaining service resilience. Furthermore, we assessed service continuity based solely on the total number of vaccine doses administered. We did not evaluate the impact of the pandemic on delayed vaccinations because the use of aggregated data limited the scope of this analysis. Delayed vaccination has been widely reported in many countries. We also did not also analyse the impact of fixed and outreach deliveries separately because the data were recorded together. Finally, we did not explore how disruptions to other essential services, such as maternal healthcare, might have impacted immunization services during the pandemic.

Conclusion

Immunization services in the Sidama region demonstrated remarkable resilience during the COVID-19 pandemic. This resilience was primarily driven by the capacity of decentralized primary healthcare services through health posts and health centers which absorbed a significant shift in demand away from hospitals, particularly during the initial lockdown period. The subsequent recovery of hospital-based services, alongside the continued strong performance of lower-level facilities, underscores the complementary role of a multi-tiered health system.

Our findings highlight that the core strength of the immunization system lies in its decentralized structure. To bolster future pandemic preparedness, health policies must prioritize (1) strengthening primary healthcare as the first line of defense, (2) formalizing linkages and task-sharing across different facility levels, and (3) implementing pre-emptive strategies, such as dynamic public communication, to minimize early disruptions at any single point of service. Future research should investigate the role of the private sector and the resilience of other essential health services to provide a holistic understanding of health system performance during crises.

Implications for policy and practice

The central role of primary healthcare facilities, especially health posts and health centers, in sustaining routine immunization highlights the need for continued investment in infrastructure, workforce, supply chains, and community-based delivery. Their resilience reinforces the importance of institutionalizing decentralized models, including task-sharing with health extension workers and stronger referral pathways, within emergency preparedness plans. Hospitals also require early targeted support to limit service disruption through revamped staffing, dedicated routine-care spaces, and strengthened infection prevention and control. Better real-time use of DHIS2 data would enable faster detection of immunization gaps and more responsive resource allocation. Collectively, these measures can enhance Ethiopia’s health system resilience and ensure continuity of routine immunization during future public health crises.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

BCG

Bacille Calmette–Guérin

DHIS₂

District Health Information System 2

GAVI

Global Alliance for Vaccines and Immunization

IPV

Inactivated Poliovirus Vaccine

LMIC

Low Middle-Income Countries

MDG

Millennium Development Goal

OPV

Oral Polio Vaccine

PCV

Pneumococcal Conjugate Vaccines

Pentavalent

DPT-Hib-HepB

SAS

Statistical Analysis System

SDG

Sustainable Development Goal

UNICEF

United Nations International Children’s Emergency Fund

WHO

World Health Organization

Author contributions

AG conceptualized the study, developed research design, facilitated data collection, finalized the first manuscript preparation. NB supported data collection, conducted the whole data analysis of first manuscript, and wrote the first draft of the manuscript. TA participated in data collection and manuscript editing. NY, MW, MA, AA, SH, and DD contributed to data collection and manuscript preparation. DH contributed to data analysis and manuscript preparation. AA also supported data analysis and manuscript preparation. AA supervised the overall study implementation and edited the first manuscript. All authors approve the first manuscript. AG and AA revised the original manuscript based on the reviewers’ comments (conducted further analysis, revised introduction, methods, result and discussion section). AA further edited the grammar of the manuscript before re-submission. All authors approved the current revised manuscript.

Funding

This research project is partly funded by Hawassa University for data collection. However, the funder had no role in the conceptualization, design, data collection, analysis, decision to publish, or manuscript preparation. The authors have no dedicated funding to cover publication cost.

Data availability

The datasets used and/or analyzed during the current study are available from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

This study was conducted in compliance with the Declaration of Helsinki (https://www.wma.net/policies-post/wma-declaration-of-helsinki/). We analysed secondary routine immunization data on vaccine doses administered before and during the COVID-19 pandemic. As data were obtained from routine vaccine administration reports submitted to DHIS2 and did not contain any identifiable patient information, informed consent was not required. We have received ethical approval and a waiver of consent from the Institutional Review Board (IRB) of Hawassa University College of Medicine and Health Sciences (Ref No. IRB/076/13–5 March 2021).

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests