Determinants of incomplete childhood vaccination among children aged 0–23 months in Ghana: a facility-based cross-sectional study

Abstract

Background

Vaccination provides lifelong protection, preventing an estimated 600,000 adult deaths and 2.5 million childhood deaths globally each year. Although Ghana has made significant progress in routine vaccination, disparities persist across regions and districts. National statistics often mask these local-level gaps. This study aimed to identify the factors influencing incomplete vaccination among children aged 0–23 months at Tetteh Quashie Memorial Hospital in Akuapem North.

Methods

A cross-sectional study was conducted among 214 caregivers of children aged 0–23 months. Kobo Collect was used as the data collection tool and the data were exported to STATA v17.0 for analysis. Vaccination status was assessed against Ghana’s Expanded Programme on Immunization (EPI) schedule. Descriptive statistics, Fisher’s exact test, and logistic regression analyses were performed to identify factors associated with incomplete vaccination, with p-values < 0.05 considered statistically significant at 95% confidence level. Results were presented in tables and graphs.

Results

The prevalence of incomplete childhood vaccination was 27.1% (95% CI [21.3%–33.6%]). Muslim mothers (aOR = 12.9; 95% CI: 4.11–18.22; p = < 0.01) and single-parent households (aOR = 13.9; 95% CI: 4.35–44.97; p < 0.01) were more likely to have incomplete vaccination, while awareness of vaccine-preventable diseases (aOR = 0.01; 95% CI: 0.00-0.02; p < 0.01) was less likely to be associated with incomplete vaccination. Children aged 5–6 months were less likely to be incompletely vaccinated (aOR = 0.29; 95% CI: 0.11–0.74; p = 0.010). Children aged 9 months and above were more likely to be incompletely vaccinated (aOR = 2.7; 95% CI: 1.24–5.99; p = 0.012). Fourth-born children were less likely to be incompletely vaccinated compared to firstborns (aOR = 0.16; 95% CI: 0.03–0.74; p = 0.020). Health-related factors, including the time from home to the vaccination facility (< 0.01) and the average waiting time at the facility (p < 0.01), were significantly associated with vaccination status.

Conclusion

This study found that vaccination coverage declined across successive doses, with incomplete vaccination associated with caregiver religion, family setting, low awareness, and the child’s age. Targeted education and improved access are needed to increase completion rates and achieve universal vaccination goals.

Introduction

Vaccination remains one of the most effective and cost-efficient public health interventions for reducing the global burden of infectious diseases. It plays a critical role in lowering morbidity and mortality, especially among children under five years of age [1]. Globally, vaccination prevents an estimated 2 to 3 million deaths annually [2, 3]. The benefits of vaccination go beyond early childhood, helping to prevent an estimated 600,000 adult deaths and an astounding 2.5 million child deaths globally [2, 4]. Despite the success achieved by global vaccination efforts, coverage has plateaued and is suboptimal in some Low- and Middle-Income Countries today [5].

The Expanded Programme on Immunization (EPI), initiated by the World Health Organization in 1974, has had a significant impact on vaccine coverage and access. The services include prevention against life-threatening conditions such as tuberculosis, poliomyelitis, diphtheria, pertussis, tetanus, Hepatitis B, and Haemophilus influenza type B infections [6, 7]. Despite these interventions, over 31 million under-five children in sub-Saharan Africa suffer from vaccine-preventable illnesses, with over half a million reported fatalities due to poor utilization of vaccines each year [8]. The failure to provide comprehensive vaccinations was responsible for half a million fatalities in sub-Saharan Africa, with 19.4 million kids worldwide not receiving the third dose of diphtheria, pertussis, and tetanus combined vaccinations in 2018, including 13.5 million not receiving even the first dose of DPT3 vaccinations [9].

Within the context of Ghana, childhood vaccination is an integral part of its overall strategy to combat under-five morbidity and mortality. The Ghana Health Service manages the EPI scheme and works to ensure that all eligible children receive vaccinations against the main causes of childhood illnesses [6]. Although Ghana has achieved significant improvement in its overall routine vaccination coverage, there still exist inequities in this area [7]. The transition initiated by this nation towards self-reliance, including the transition away from GAVI support, has revealed weaknesses in its vaccination systems, thereby sometimes resulting in lower vaccination coverage [10, 11].

National-level data often conceal gaps at the regional and district level. For instance, a study reported a relatively low prevalence of incomplete vaccination (15.5%) among children aged 12–23 months, suggesting generally good vaccine uptake. Nevertheless, coverage of the second dose of measles vaccines at 18 months was found to be alarmingly low at only 23.9% [12]. Another case in point can be seen in the Volta Region, Ghana, at the Volta Region, specifically at the Hohoe Municipality, where in facility-based research among 378 caregivers of those aged 0–23 months, the prevalence of Missed Opportunities for Vaccination (MOV) was found to be 22.2% [8].

Children who are not up to date with all recommended vaccinations by 23 months are referred to as incompletely vaccinated, thereby increasing their predisposition to infections, serious long-term health conditions, and preventable fatalities. Various research efforts across sub-Saharan Africa, including Ghana, have identified significant socio-demographic and facility-related factors contributing to incomplete vaccination coverage [13]. These factors include maternal level of education, caregiver’s level of awareness and perception, income, distance to health facilities, birth order, and the child’s age [14, 15]. The majority of these research analyses, however, were conducted at the national or regional levels. There is limited evidence from facility-based studies focusing on determinants of incomplete vaccination among children aged 0–23 months in semi-urban Ghana. This research, however, sought to determine the extent of incomplete childhood vaccination among those aged 0–23 months at the facility level in Akuapem North, Ghana.

Methods

Study design

A facility-based cross-sectional study design was used to estimate population parameters, such as the proportion of incomplete vaccination, the association between child characteristics and incomplete childhood vaccination, and factors associated with incomplete childhood vaccination [16]. The research design enables the collection and statistical analysis of data to identify patterns and correlations in participants’ responses.

Study site description

The study was conducted at Tetteh Quarshie Memorial Hospital, the only public hospital within a 30-kilometer radius that provides healthcare services to over 120,000 residents of the semi-urban Akuapem Ridge. The hospital offers both outpatient and inpatient services, with the outpatient department providing general medical care for patients with various health conditions. The inpatient department has well-equipped wards where patients can receive specialized medical care for conditions requiring hospitalization. The hospital also has a dedicated maternity ward where expectant mothers can receive antenatal, delivery, and postnatal care. The maternity ward is staffed by trained midwives and obstetricians to ensure safe delivery and postnatal care for mothers and their newborns.

Study population

The study population consisted of nursing mothers with children aged 0–23 months who sought healthcare at the Tetteh Quarshie Memorial Hospital.

Sample procedure and techniques

To determine the sample size for this study, the Cochran formula was used. The specific formula utilized was:

Where n= sample size to be determined, Z = Z-score (1.96 at 95% confidence interval), P= proportion (Estimated prevalence of incomplete vaccination 15.5% [17]), d= precision or margin of error 5% (0.05)

Thus, the minimum required sample size was approximately 201 participants. To account for potential non-responses and incomplete data, an additional 6% was added to the minimum sample size, resulting in a final sample size of 214 participants.

Participants were selected using purposive sampling because the study focused on nursing mothers with children aged 0–23 months attending Tetteh Quarshie Memorial Hospital, and probability sampling was not feasible due to financial and time constraints. Although this approach limits representativeness, it ensured the inclusion of participants most relevant to the study objectives. Individuals who met the inclusion criteria were identified and approached based on their relevance to the study objectives. Eligible participants were informed about the study and invited to provide consent before participation. In cases where an eligible individual declined to participate, another suitable participant who met the inclusion criteria was purposively selected until the required sample size was achieved, while accounting for non-responses during data collection.

Data collection tool

Quantitative data were collected through face-to-face interviews conducted by the principal investigators using a structured questionnaire. The interviews were administered directly to eligible participants to ensure clarity and completeness of responses. Data were collected digitally using the Kobo Collect to enhance accuracy and efficiency.

Study variables

The outcome variable was vaccination status, where incomplete vaccination is defined as missing at least one age-appropriate EPI vaccine by 23 months, measured by vaccine uptake (OPV, PENTA, IPV, PCV, MR, Men A, and Yellow Fever). The predictor variables consisted of participants’ socio-demographic characteristics (age, education level, marital status, religion, parity, occupation, family setting, and awareness of vaccine-preventable diseases) and health service-related factors (travel time to the vaccination facility, means of transport to the facility, and average waiting time at the facility).

Data analysis

Data collected through Kobo Collect were thoroughly checked for completeness and subsequently analysed using Stata version 17. All variables were appropriately coded numerically (binary responses were coded as 0 and 1, while multi-category variables were assigned values such as 1, 2, and 3). Descriptive statistics were used to summarize the data, with frequencies and percentages presented for categorical variables (demographic characteristics) and means with standard deviations for continuous variables (caregiver age). Fisher’s exact test and logistic regression were used to examine preliminary associations between vaccination status and key explanatory variables including caregivers’ demographics, children’s characteristics, and healthcare-related factors. Factors found to be statistically significant (p < 0.05) were included in a multivariable logistic regression model to identify independent predictors of incomplete vaccination. The Hosmer-Lemeshow test for vaccination status (p = 0.92) indicated no significant deviation between observed and predicted values. The Variance Inflation Factor (VIF) results show no serious multicollinearity among the sociodemographic variables. All VIF values are below 2, and tolerances are well above 0.2, indicating that none of the variables are excessively correlated. The mean VIF of 1.36 further confirms that multicollinearity is minimal and the variables can be reliably included in the regression model (see Table 1). Additionally, caregiver-reported reasons for incomplete vaccination were thematically categorized and presented as proportions.

Table 1.

Multicollinearity diagnostics (VIF) for independent variables

Independent Variable	VIF	Tolerance (1/VIF)
Age	1.48	0.67
Educational Level	1.22	0.82
Marital Status	1.87	0.54
Religion	1.02	0.98
Parity of the mother	1.38	0.72
Occupational Status	1.28	0.78
Family setting	1.46	0.68
Aware of diseases vaccines	1.15	0.87
Mean VIF	1.36	__

Ethics approval and consent to participate

Ethical approval for this study was obtained from the University of Health and Allied Sciences Research Ethics Committee (UHAS-REC) under reference number UHAS-REC A.4[115]23–24. Additional administrative permission was granted by the Medical Superintendent of Tetteh Quarshie Memorial Hospital, where the study was conducted. Parents or legal guardians were fully informed about the study’s objectives and its voluntary nature. Verbal informed consent was obtained from the parents or legal guardians of children prior to participation, as most participants were not literate. Participants were assured that their decision to participate or withdraw at any time would have no negative consequences.

To protect participant privacy, minimal personally identifiable information was collected. Confidentiality was maintained by assigning unique codes to each participant and storing data on a password-protected server accessible only to the research team. Interviews were conducted privately, and all responses were handled in accordance with the ethical guidelines outlined in the Declaration of Helsinki (2024), ensuring the dignity, safety, and rights of all participants.

Results

Socio-demographic characteristics of caregivers

Table 2 presents the demographic characteristics of the caregivers, all of whom were the children’s mothers. Nearly half 104 (48.6%) were below 30 years of age, with a mean age of 29.4 years (± 6.7). The majority, 155 (72.4%), had attained at least secondary education, followed by Tertiary 46 (21.5%). Regarding marital status, 134 (62.6%) were married, and 80 (37.4%) were single. Most of the respondents, 210 (98.1%), were identified as Christians. In terms of parity, approximately 117 (55%) had between two and four children, and 83 (38.32%) had one child. More than half, 135 (63.1%), were traders, while less than one-fifth, 33 (15.4%), were unemployed. The majority, 179 (83.6), were from a two-parent household. Notably, all caregivers 214, 100%) resided in urban areas.

Table 2.

Socio-demographic factors of caregivers

Variable	Frequency (=214)	Percentage (%)
Care-givers relationship with the child
Mother	214	100.0
Age mean (±SD)	29.4 (±6.7)
<30	104	48.6
30–39	96	44.9
40+	14	6.5
Highest level attended
None	3	1.4
Primary	10	4.7
Secondary	155	72.4
Tertiary	46	21.5
Mothers’ current marital status
Married	134	62.6
Single	80	37.4
Current religion
Christian	210	98.1
Muslim	4	1.9
Parity of the mother
1 child	83	38.3
2–4 children	117	54.7
5 or more	15	7.0
Present occupational
Unemployed	33	15.4
Civil servant	46	21.5
Trader	135	63.1
Family setting
Two-parent households	179	83.6
Single parent household	35	16.4
Area of residence
Urban	214	100.0

Child characteristics

Table 3 presents the characteristics of the children and their vaccine uptake records. Approximately 76 (36%) of the children were less than three months old, with mean age of 5.8 (± 4.7). Almost two-fifth 84 (39%), were ranked “1” on the child’s birth order. More than half, 113 (53%), of the children were males. The majority, 203 (95%) of the children had both of their parents alive.

Table 3.

Child characteristics and vaccine uptake

Variable	Frequency (=214)	Percentage (%)
Child’s age (months)	5.8 (±4.7)
<3	76	35.5
3–4	15	7.0
5–6	62	28.9
7–8	15	7.0
9+	46	21.6
The rank or birth order the child
1	84	39.3
2	62	28.9
3	38	17.8
4 and above	30	14.0
Sex of the child
Female	99	46.7
Male	113	53.3
Are both parents alive
No	11	5.1
Yes	203	94.9

Dose-specific vaccination prevalence

Regarding the vaccine uptake, individuals who took BCG antigen were 93.5%, OPV 0 (94.4%), OPV 1 (62.2%), OPV 2 (43.9%), OPV 3 (34.6%), PENTA 1 (62.1%), PENTA 2 (43.9%), PENTA 3 (33.2%), IPV (33.2%), PCV 1 (60.2%), PCV 2 (43.0%) PCV 3 (32.7%), MR 1 (12.1%), MR 2 (0,5%), Men A (1.4%) and YF (10.7%) as seen in (Table 4).

Table 4.

Dose-specific Vaccination Prevalence

Eligible Age	Vaccine Offered (Dose)	Eligible	Vaccinated (%)	Not vaccinated (%)
Birth	BCG	214	200 (93.5)	14 (6.5)
	OPV (O)	214	202 (94.4)	12 (5.6)
6 Weeks	OPV (1)	163	133 (62.1)	30 (37.9)
	PENTA 1	163	133 (62.1)	30 (37.9)
	PCV 1	163	129 (60.3)	34 (39.7)
10 Weeks	OPV 2	137	94 (43.9)	43 (56.1)
	PENTA 2	137	94 (43.9)	43 (56.1)
	PCV 2	137	92 (43.0)	45 (57.0)
14 Weeks	OPV 3	126	74 (34.6)	52 (65.4)
	PENTA 3	126	71 (33.2)	55 (66.8)
	IPV	126	71 (33.2)	55 (66.8)
	PCV 3	126	70 (32.7)	56 (67.3)
9 Months	MR 1	46	26 (56.5)	20 (43.5)
	YF	46	23 (50.0)	23 (50.0)
18 Months	MR 2	7	1 (14.3)	6 (85.7)
	MEN-A	7	3 (42.9)	4 (57.1)

BCG Bacille Calmette-Guérin vaccine, OPV (O) Oral polio vaccine zero dose, OPV (1) Oral polio vaccine first dose, PENTA 1 Pentavalent vaccine first dose (containing diphtheria, tetanus, pertussis, hepatitis B, and Haemophilus influenzae type b vaccines), PCV 1 Pneumococcal conjugate vaccine first dose, OPV 2 Oral polio vaccine second dose, PENTA 2 Pentavalent vaccine second dose, PCV 2 Pneumococcal conjugate vaccine second dose, OPV 3 Oral polio vaccine third dose, PENTA 3 Pentavalent vaccine third dose, IPV Inactivated polio vaccine, PCV 3 Pneumococcal conjugate vaccine third dose, MR 1 Measles-rubella vaccine first dose, YF Yellow fever vaccine, MR 2 Measles-rubella vaccine second dose, MEN-A Meningococcal A vaccine

Vaccination status among children

Figure 1 shows the vaccination status of children in the study. The results showed that 58 (27.1%) (95% CI [21.3%–33.6%]) were incompletely vaccinated, while 156 (72.9%) (95% CI [66.4%–78.7%]) were completely vaccinated. Figure 2 also shows the reasons for incomplete vaccination, where vaccine stockout resulted in (12.1%) and 2.4% said they were not aware of the date of vaccination.

Fig. 1.

Vaccination status among children aged 0–23 months attending Tetteh Quarshie Memorial Hospital

Fig. 2.

Reported reasons for incomplete childhood vaccination among caregivers

Healthcare-related factors

Table 5 shows that 96.3% of caregivers spent between 30 and 59 min traveling from their homes to the vaccination facility. The majority (95.8%) used a vehicle as a means of transport to the facility. About 97% spent an average of 30 to 59 min waiting to be seen at the facility. None of the caregivers had ever misplaced their child’s infant welfare care.

Table 5.

Health care related factors

Variable	Frequency (n=214)	Percentage (%)
Minutes taken from home to the vaccination facility
30–59 mins	206	96.3
Less than 30 mins	8	3.7
Means of getting to the health facility
Vehicle	205	95.8
Walking	9	4.2
Average waiting time at the facility
30–59 mins	207	96.7
Less than 30 mins	7	3.3
Ever replaced the child infant welfare card
No	214	100.0

Logistic regression analysis of determinants of vaccination status among children aged 0–23 months

Table 6 shows that caregiver religion (0.030), family setting (0.015), and awareness (< 0.01) were significantly associated with incomplete childhood vaccination. After controlling for potential confounders, Muslim were 13 times more likely of having an incompletely vaccinated child compared to Christian mothers (aOR = 12.9; 95% CI: 4.11–18.22; p = < 0.01). Children in single-parent households were also 13.9 times more likely to be incompletely vaccinated than those in two-parent households (aOR = 13.9; 95% CI: 4.35–44.97; p < 0.01). Caregivers who were aware of vaccination had lower odds of incomplete vaccination than those who lacked awareness (aOR = 0.01; 95% CI: 0.00–0.02.00.02; p < 0.01).

Table 6.

Association between incomplete vaccination status and caregiver’s demographic factors

Covariates	Vaccination status		Fisher Exact Stat.	cOR (95% CI), p-value	aOR (95% CI), p-value
	Complete 156 (72.9%)	Incomplete 58 (27.1%)
Age			0.294
<30	71 (68.3)	33 (31.7)		Ref
30–39	75 (78.1)	21 (21.9)		0.60 (0.32–1.14.32.14), 0.118
40+	10 (71.4)	4 (28.6)		0.86 (0.25–2.95.25.95), 0.811
Highest level attended			0.122
None	2 (100)	0 (0)		Ref
Primary	9 (90.0)	1 (10.0)		5.79 (0.65–49.63.65.63), 0.109
Secondary	115 (75.0)	39 (25.0)		1.9 (0.96–3.86.96.86), 0.064
Tertiary	28 (60.9)	18 (39.1)		6.92 (0.72–1.23.72.23), 0.084
Mothers’ current marital status			0.081
Single	64 (80.0)	16 (20.0)		Ref
Married	92 (68.7)	42 (31.3)		0.55 (0.28–1.06.28.06), 0.073
Current religion			0.030
Christian	155 (73.8)	55 (26.2)		Ref	Ref
Muslim	1 (25.0)	3 (75.0)		0.12 (0.12–1.16.12.16),0.067	12.9 (4.11–18.22.11.22), <0.01
Parity of the mother			0.184
1 child	58 (70.7)	24 (29.3)		Ref
2–4 children	84 (71.8)	33 (28.2)		0.95 (0.51–1.77.51.77), 0.870
5 or more	14 (93.3)	1 (6.7)		.017 (0.02–1.39.02.39), 0.098
Present occupational			0.108
Unemployed	21 (63.6)	12 (36.4)		Ref
Civil servant	30 (65.2)	16 (34.8)		0.93 (0.37–2.37.37.37), 0.885
Trader	105 (77.8)	30 (22.2)		0.5 (0.37–2.37.37.37), 0.096
Family setting			0.015
Two-parent households	137 (76.1)	43 (23.9)		Ref	Ref
Single-parent household	19 (55.9)	15 (44.1)		2.51 (1.18–5.37.18.37), 0.017	13.9 (4.35–44.97.35.97), <0.001
Awareness of vaccine-preventable diseases			0.000
No	6 (12.8)	41 (87.2)		Ref	Ref
Yes	150 (89.8)	17 (10.2)		0.02 (0.00–0.04.00.04), <0.01	0.01 (0.00–0.02.00.02), <0.01

The bolded numbers indicate statistically significant results

Table 7 shows that Child’s age (< 0.01) and rank of the child (0.030) were significantly associated with vaccination status. After controlling for potential confounders, children aged 5–6 months were less likely to be incompletely vaccinated than those under 3 months (aOR = 0.29; 95% CI: 0.11–0.74; p = 0.010). Children nine months old and above were nearly three times more likely to be incompletely vaccinated (aOR = 2.7; 95% CI: 1.24–5.99; p = 0.012). Fourth-born children and above were less likely to be incompletely vaccinated compared to firstborns (aOR = 0.16; 95% CI: 0.03–0.74; p = 0.020).

Table 7.

Association between incomplete vaccination status and child’s characteristics

Covariates	Vaccination status		Fisher Exact Stat.	cOR (95% CI), p-value	aOR (95% CI), p-value
	Complete 156 (72.9%)	Incomplete 58 (27.1%)
Childs age (months)			<0.001
<3	53 (69.7)	23 (30.3)		Ref	Ref
3–4	14 (93.3)	1 (6.7)		0.16 (0.02–1.33.02.33),0.090	0.17 (0.20–1.35.20.35), 0.093
5–6	55(88.7)	7(11.3)		0.29 (0.12–0.74.12.74), <0.01	0.29 (0.11–0.74.11.74), 0.010
7–8	12(80.0)	3(20.0)		0.58 (0.15–2.24.15.24), 0.426	0.69 (0.17–2.79.17.79), 0.606
9+	24(52.2)	22(47.83)		2.5 (1.18–5.36.18.36), 0.017	2.7 (1.24–5.99.24.99), 0.012
The rank or birth order the child			0.030
1	60 (71.4)	24 (286)		Ref	Ref
2	43 (69.4)	19 (30.7)		1.10 (0.54–2.27.54.27), 0.786	0.82 (0.37–1.78.37.78), 0.612
3	25 (65.8)	13 (34.2)		1.3 (0.57–2.95.57.95), 0.531	0.96 (0.39–2.32.39.32), 0.923
14 and above	28 (93.3)	2 (6.7)		0.18 (0.39–0.81.39.81), 0.025	0.16 (0.03–0.74.03.74), 0.020
Sex of the child			0.357
Female	69 (69.7)	30 (30.3)		Ref
Male	87 (75.7)	28 (24.4)		0.74 (0.40–1.35.40.35), 0.329
Both parents being alive			0.065
No	6 (54.5)	5 (45.5)
Yes	117 (57.6)	86 (42.4)

The bolded numbers indicate statistically significant results

Some adjusted estimates were accompanied by wide confidence intervals, particularly for categories with small sample sizes (e.g., religion and higher birth order groups). These wide intervals suggest limited precision of the estimates, likely due to sparse data within certain subgroups. The magnitude of these associations should therefore be interpreted cautiously despite statistical significance.

Table 8 shows the association between incomplete vaccination status and healthcare-related factors. The results revealed significant associations between the minutes taken from home to the vaccination facility (< 0.01) and the average waiting time at the facility (< 0.01) with vaccination status.

Table 8.

Association between incomplete vaccination status and healthcare-related factors

Covariates	Vaccination status		Fisher Exact Stat.
	Incomplete 58 (27.1%)	Complete 156 (72.9%)
Minutes taken from home to the vaccination facility			0.000
Less than 30 mins	8 (100)	0 (0)
30–59 mins	50 (24.3)	156 (75.7)
Means of getting to the health facility			0.063
Vehicle	53 (25.9)	152 (74.2)
Walking	5 (55.6)	4 (44.4)
Average waiting time at the facility			0.000
Less than 30 mins	7 (100)	0 (0.0)
30–59 mins	51 (24.6)	156 (75.4)

The bolded numbers indicate statistically significant results

Discussion

The study identified determinants of incomplete childhood vaccination among children aged 0–23 months in Akuapem North, Ghana, and the factors identified were the child’s age, birth order, caregivers’ knowledge, health system and family setting. The findings of this study reveal a low prevalence of incomplete childhood vaccination among children aged 0–23 months in Akuapem North, Ghana. This rate is slightly higher than the national average reported in previous studies, such as the 15.5% prevalence observed among children aged 12–23 months in Ghana [12]. The study revealed a significant decline in vaccination coverage as children grew older. The sharp drop in dose-specific coverage between early antigens is due to vaccine stockouts and lack of awareness of scheduled vaccination dates as mentioned by participants. A similar study conducted in Peru observed that vaccination rates dropped as the child’s scheduled age increased, thus coverage for vaccines scheduled at 7–18 months was much lower than for those at 0–2 months [18]. Also, children aged 9 months or older were more likely to be incompletely vaccinated than those under 3 months. These findings support the “caregiver fatigue” hypothesis, where adherence wanes over time due to competing priorities, reduced perceived urgency for later doses, vaccine stock-outs, or logistical challenges in returning to health facilities [19]. Similar patterns have been documented in other sub-Saharan African settings, where later doses of multiantigen vaccines such as PENTA 3 (33.2% coverage) and PCV 3 (32.7% coverage) consistently show lower uptake than initial doses [15, 20]. The findings suggest that vaccination gaps are more likely to occur at later ages, potentially shifting disease risk to older infants and toddlers and increasing the risk of outbreaks of vaccine-preventable diseases. From a public health perspective, this underscores the need to adapt vaccination programmes to place greater emphasis on continuity of care, timely completion of doses, and targeted support for caregivers as children progress through the vaccination schedule.

Birth order also emerged as a critical factor, with fourth children less likely to be incompletely vaccinated compared to firstborns. This may be due to increased maternal experience and familiarity with vaccination schedules gained from caring for older children. Notably, this finding contrasts with studies from 85 low- and middle-income countries, where birth order was not a significant predictor [21], suggesting contextual variations that warrant further exploration. This finding suggests that maternal experience plays an important role in ensuring complete vaccination, with higher-order births benefiting from caregivers’ prior knowledge and confidence in following vaccination schedules. This implies that first-time mothers may require additional guidance and support to reduce the risk of incomplete vaccination among firstborn children and to improve overall childhood vaccination coverage.

Contrary to expectations, children from Muslim religious backgrounds and single-parent households were more likely to be incompletely vaccinated. This unexpected association may reflect the fact that single-parent households face greater time and resource constraints, while cultural or religious factors in some Muslim communities may influence healthcare-seeking behaviors, including vaccination uptake. This finding highlights how social and economic circumstances, rather than caregiver negligence, can contribute to unequal health outcomes among children. This study aligns with broader literature linking single-parent households to poorer health outcomes [22]. Similarly, a study conducted in Japan found that Unmarried mothers were more likely to have had a childhood without vaccination than married mothers [23]. A study conducted among Muslim parents in Thailand revealed that misconceptions about vaccine safety, concerns over religious compliance, and reliance on informal community information were identified as barriers, while trust in vaccines and guidance from religious leaders promoted uptake. These findings suggest that cultural and religious factors can significantly affect vaccination decisions [24]. Our findings revealed that caregiver knowledge of vaccine-preventable diseases was low. Caregivers who were aware of vaccinations had lower odds of incomplete vaccination than those who were unaware. This suggests that awareness plays a protective role in improving vaccination completion. However, awareness alone may not be sufficient to guarantee full adherence to vaccination schedules. Dropout rates observed for later doses indicate that even informed caregivers may face practical or systemic barriers that hinder completion. Vaccination programs should extend beyond information dissemination to incorporate strategies that strengthen adherence, including structured reminder and recall systems, community-based follow-up, and improved service accessibility. Integrating targeted health education with supportive delivery mechanisms is essential to translate awareness into sustained vaccination completion and to optimize program effectiveness. Similarly, a study conducted in southwestern Nigeria revealed that caregivers with positive perceptions of vaccination were twice as likely to have their children immunized [25]. This consistency suggests that favourable attitudes and trust in vaccination programs play a critical role in improving vaccination uptake across settings. The findings highlight the importance of health education strategies that not only increase knowledge but also actively foster positive perceptions and confidence in vaccination services, as these attitudinal components appear to reinforce completion of vaccination schedules.

Surprisingly, our study revealed that logistical challenges, such as travel time and waiting times at healthcare facilities, were significantly associated with incomplete vaccination. This suggests that even in urban areas with relatively uniform access, delays, longer travel and waiting times can still hinder timely vaccination schedules. However, vaccine stockouts were reported by 12.1% of caregivers, highlighting supply chain vulnerabilities that disrupt vaccination schedules. When health facilities experience vaccine stockouts, scheduled vaccinations are missed, leading to delays in administering the required doses within the recommended timelines [26]. Such disruptions mostly arise from supply chain inefficiencies, inaccurate forecasting, inadequate cold chain infrastructure, or logistical distribution challenges, particularly in rural or underserved areas [27]. This systemic issue mirrors findings from Tanzania [28] and underscores the need for robust supply chain management to ensure consistent vaccine availability. While rescheduling mothers’ appointments to periods when vaccines are available is a common coping strategy, its effectiveness is limited, as not all return due to competing responsibilities, transport barriers, financial constraints, or reduced trust in the health system [29, 30]. The study’s results diverge from some regional findings, such as the lack of association between maternal age and vaccination status, which contrasts with studies in Senegal [31]. The high dropout rates for later doses (e.g., MR2 coverage at 0.5%) align with Ghana’s national challenges in sustaining vaccination momentum beyond infancy [32].

Limitations of the study

The cross-sectional design limits the ability to establish causality, and conducting the study in a single hospital restricts generalizability to other settings. Associations were therefore interpreted cautiously, without implying causal relationships. The small number of participants in certain subgroups (e.g., religion and eligibility for MR2 and MenA vaccines) limited the precision of effect estimates and constrained the ability to draw definitive conclusions for these categories. The use of purposive sampling could introduce selection bias; however, this risk was minimized by applying clear, predefined inclusion criteria and enrolling all eligible participants during the study period. While every effort was made to ensure participant anonymity and comfort during interviews, recall and residual biases cannot be completely ruled out. Despite these limitations, the study utilized standardized procedures and validated tools, thereby ensuring reasonable internal validity and the reliability of the findings.

Conclusion and recommendations

This study shows that although early childhood vaccination coverage is high, a substantial drop-out occurs at later doses, underscoring the need for sustained engagement with caregivers throughout the vaccination schedule. Strengthening reminder and recall systems, including appointment cards and mobile phone-based reminders, could improve adherence to scheduled visits, while routine defaulter tracing by community health workers may help identify and follow up on children who miss doses. Additionally, reinforcing vaccine supply chain management to prevent stockouts is critical to ensuring that caregivers who are present for services are not turned away. Embedding these strategies within routine maternal and child health services may enhance vaccination completion and support progress toward childhood vaccination coverage.

Abbreviations

MOV

Missed Opportunities for Vaccination

BCG

Bacilli Calmette-Guerin

DPT

Diphtheria-Pertussis-Tetanus

HPV

Human Papilloma Virus

MR

Measles-Rubella

NIP

National Immunization Program

OPV

Oral Polio Vaccine

PENTA

Pentavalent

TT

Tetanus Toxoid

Authors’ contributions

LAD and MKA conceptualized the study and designed the methodology. LAD and CA handled data collection and analysis. Data analysis was conducted by CA and DTN. DNT and CA prepared the initial draft of the manuscript. LAD, CA, DNT, MKA, IEA, FMS, and AFD reviewed the data, contributed to manuscript editing, and provided supervision. All authors approved the final version for publication.

Funding

The authors declare that they have no conflicts of interest.

Data availability

The data will be made available upon request via the corresponding author at [Christopherayisah128@gmail.com](mailto: Christopherayisah128@gmail.com).

Declarations

Ethics approval and consent to participate

This study received ethical approval from the Research Ethics Committee of the University of Health and Allied Sciences (UHAS-REC A.4[115]23–24). In line with the Declaration of Helsinki, all parents or legal guardians gave voluntary informed consent after being informed of the study’s purpose, their right to withdraw at any time without penalty, and the confidentiality of the information collected.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.