Association of knowledge and observed skills with self-reported readiness of nurses and midwives to perform cardiopulmonary resuscitation in a tertiary hospital in Ghana: a cross-sectional study

Abstract

Background

Cardiopulmonary resuscitation (CPR) is a critical emergency intervention, yet evidence on CPR competence and readiness among nurses and midwives in Ghana is limited. This study assessed CPR knowledge, observed skills, self-reported readiness, and factors associated with readiness in a tertiary hospital.

Methods

An analytical cross-sectional study was conducted among nurses and midwives in critical and emergency care units. Of 331 eligible staff, 271 participated (81.9% response rate). Knowledge was assessed using a structured test, skills were observed on a manikin using a 12-item checklist and self-reported readiness using an 11-item Likert scale. Data were analysed using descriptive statistics, chi-square tests and binomial logistic regression.

Results

Most respondents had low CPR knowledge (94.8%), while 65.7% demonstrated adequate observed skills and 90.0% reported moderate readiness. In the binomial logistic regression, adequate CPR skills were associated with moderate readiness (AOR = 12.84, 95% CI [4.09, 40.36], p < .001). Moderate knowledge was associated with lower odds of moderate readiness compared with low knowledge (AOR = 0.11, 95% CI [0.03, 0.46], p = .003), although this finding should be interpreted cautiously. Other demographic and training variables were not significant predictors.

Conclusion

CPR preparedness was characterised by low theoretical knowledge, comparatively better but incomplete observed skills and moderate self-reported readiness among the nurses and midwives. Observed CPR skills were strongly associated with readiness. Therefore, regular simulation-based training and competency assessment combining knowledge testing with observed practical performance are recommended.

Clinical trial number

Not applicable.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s12912-026-05327-y.

Introduction

Cardiac arrest is a major cause of mortality globally and requires immediate intervention to improve survival outcomes [1]. Cardiopulmonary resuscitation (CPR) is a critical life-saving procedure used to restore spontaneous circulation and preserve organ perfusion through chest compressions and assisted ventilation [2, 3]. Early initiation of high-quality CPR has been shown to significantly increase survival rates and reduce neurological complications among cardiac arrest victims. However, despite standardised international resuscitation guidelines, survival following in-hospital cardiac arrest is suboptimal, partly because successful resuscitation depends on healthcare providers being able and prepared to initiate high-quality CPR promptly [4–6].

The challenges associated with CPR implementation are particularly evident in low- and middle-income countries (LMICs), where limited training opportunities, inadequate equipment, irregular refresher training and restricted access to simulation-based learning can compromise resuscitation performance [7, 8]. Evidence from studies conducted in Africa similarly identifies poor retention of psychomotor CPR skills, inadequate emergency equipment and weaknesses in coordinated emergency response [9, 10]. These limitations can delay CPR initiation and reduce adherence to recommended resuscitation standards, thereby contributing to poor cardiac arrest outcomes [11].

In Ghana, tertiary hospitals function as referral centres responsible for managing critically ill patients and emergencies such as cardiac arrest. Nurses and midwives constitute an important component of the frontline clinical workforce and may be among the first healthcare professionals to identify cardiac arrest and initiate resuscitation within their respective clinical areas. Their competence and preparedness to perform CPR therefore have important implications for timely emergency response and patient outcomes. Moreover, nurses and midwives worked within critical and emergency care units, were expected to respond to cardiopulmonary emergencies under the same institutional resuscitation procedures and shared responsibility for initiating basic CPR when indicated. Therefore, differences in professions were not the focus of the study.

Existing evidence from Ghana suggests variations in CPR knowledge and practical competence among nurses and other healthcare professionals [12], which have been associated with limited opportunities for simulation, inadequate refresher training, and inconsistent exposure to standardized resuscitation training [13]. Similarly, previous studies in Ghana and other African settings have examined healthcare providers’ knowledge, attitudes and reported CPR practices while little attention has been given to their readiness to perform CPR. Therefore, as guided by the Social Cognitive Theory, which emphasise the interaction between knowledge, behavioural capability, self-efficacy and action, readiness in this study was operationalised as healthcare providers’ self-reported preparedness to initiate and participate in CPR when confronted with a cardiac arrest. It therefore reflects perceived confidence, willingness, and preparedness to act, whereas CPR skills represent objectively observed psychomotor performance during CPR on a manikin. This distinction is important because possessing factual knowledge or demonstrating a skill under assessment conditions does not necessarily mean that a healthcare professional perceives themselves as ready to initiate CPR during an actual emergency. Given the time-sensitive nature of cardiac arrest, delays or hesitation in initiating resuscitation may adversely affect survival.

The research gap addressed by this study is the limited evidence regarding whether CPR knowledge and observed CPR skills are associated with nurses’ and midwives’ self-reported readiness to perform CPR in tertiary healthcare settings in Ghana. This lack of evidence may constrain efforts to design targeted interventions for strengthening emergency response systems, workforce preparedness, and resuscitation outcomes. Thus, understanding these relationships are important because training programmes that focus exclusively on knowledge acquisition may not adequately address deficiencies in psychomotor competence or preparedness to respond during actual cardiac arrest events. Accordingly, the primary objective of this study was to determine the association of CPR knowledge and observed CPR skills on the self-reported readiness of nurses and midwives to perform CPR in a tertiary hospital in Ghana. The findings are expected to inform institutional training policies, strengthen resuscitation preparedness, and contribute to efforts to reduce avoidable mortality from cardiac arrest. The study also aligns with Sustainable Development Goal Target 3.4, which focuses on reducing premature mortality from non-communicable diseases, and Target 3.C, which emphasises strengthening the health workforce in developing countries.

Methods

Research design and population

An analytical cross-sectional design was used for the study. The study was conducted at a tertiary hospital in Ghana. For confidentiality and to preserve institutional anonymity, the name of the hospital has been omitted from the manuscript. The hospital provides specialist and emergency services to patients referred from other health facilities and neighbouring regions. The study population consisted of nurses and midwives working within seven selected critical and emergency care units of the hospital comprising the Delivery Suite, Obstetrics and Gynaecology Unit, Intensive Care Unit, Surgical Theatre Recovery, Paediatric Medical Ward, Neonatal Intensive Care Unit and Accident and Emergency Department. These units were selected because nurses and midwives working within them were considered more likely to encounter patients requiring CPR.

The sampling frame comprised 331 nurses and midwives after applying exclusion criteria. Eligible participants were nurses and midwives who had at least one year of professional work experience, were working in one of the selected units and were involved in direct patient care. Student nurses and midwives and eligible staff who were on leave during the data collection period were excluded. Since recruitment was intended as a census of the eligible population, no sample-size calculation was used to restrict recruitment, therefore, all eligible nurses and midwives were invited to participate.

Sampling procedure

A census approach was used in which all eligible nurses and midwives in the selected clinical units were invited to participate, thereby maximising coverage of the accessible population. The target population was drawn from the Delivery Suite, Obstetrics and Gynaecology Unit, Intensive Care Unit, Surgical Theatre Recovery, Paediatric Medical Ward, Neonatal Intensive Care Unit, and Accident and Emergency Department. These units were selected because staff routinely cared for patients at risk of clinical deterioration or cardiac arrest and were therefore considered more likely to encounter situations requiring CPR. Consequently, the target population represented nurses and midwives working in selected critical and emergency care units rather than all nurses and midwives in the tertiary healthcare facility. Of the 331 nurses and midwives in the sampling frame, 271 provided complete data giving a response rate of 81.9%. The remaining 60 nurses and midwives, 46 declined participation and 14 provided incomplete data and was not added to the analysis. Consequently, the data of 271 respondents were used for the final analysis. Consequently, the target population represented nurses and midwives working in selected critical and emergency care units rather than all nurses and midwives in the hospital or in Ghana. The data collection was commenced on 1st April to 29th May, 2024.

Measures

The data collection instrument was developed from related literature and consisted of sections assessing sociodemographic characteristics, CPR knowledge, observed CPR skills, and self-reported readiness to perform CPR. Demographic characteristics were assessed using seven items, including sex and age. Knowledge of CPR was measured using 22 dichotomous items adapted from Rajeswaren et al. [14]. Examples included: “CPR training and retraining is necessary for nurses and midwives” and “CPR should be conducted on the patient immediately before informing the doctor.” For scoring, each response that corresponded with the prespecified correct answer was coded 1 and each incorrect response was coded 0, producing a possible total knowledge score of 0–22 (see Supplementary File).

Although the internal consistency of the CPR knowledge items was relatively low (KR-20 = 0.475), the items were retained because the knowledge assessment was designed as a criterion-referenced measure sampling distinct aspects of CPR knowledge rather than as a unidimensional psychometric scale. The content covered core resuscitation domains reflected in international guidance, including recognition and initiation of cardiopulmonary resuscitation, pulse assessment, airway and ventilation management, chest-compression technique, compression rate and depth, compression-to-ventilation ratios, and paediatric basic life support. Accordingly, the low KR-20 was interpreted as evidence of limited internal consistency and is acknowledged as a limitation. However, the content validity of the instrument was supported through expert review. The experts assessed the items for their relevance, clarity, appropriateness, comprehensiveness and alignment with the study objectives and established CPR guidelines.

CPR skills were measured using 11 dichotomous observational items adapted from Umran and Dilek [15]. Examples included “Ensure safety of the environment,” “Confirm that the patient is not breathing and/or is pulseless,” and “Call for help.” Skills were assessed objectively by observing each participant perform a standardised CPR scenario on a manikin. Each correctly demonstrated action was coded 1 and each incorrectly performed or omitted action was coded 0, producing a possible skill score of 0–11 (see Supplementary File). Each participant completed a standardised infant CPR simulation using a manikin. The manikin provided real-time feedback on CPR parameters such as compression rate, compression depth, chest recoil and ventilation. All participants were assessed using the same predetermined scenario and standard instructions to ensure consistency across assessments. The assessments were conducted by two trained research assistants who received prior orientation and practical training on the use of the observational checklist, the simulation protocol, scoring criteria and interpretation of each skill item. Before data collection, the assessors underwent assessment procedure to ensure consistency in scoring. Each participant was independently rated by both assessors and interrater agreement was evaluated using Cohen’s kappa and scored 0.82 which is considered almost perfect. The skills scale demonstrated good reliability (α = 0.843).

Self-reported readiness to perform CPR was assessed using 12 items rated on a five-point Likert scale ranging from 1 = Strongly Disagree to 5 = Strongly Agree, adapted from Alwidyan et al. [16]. Example statements included “I am confident in my ability to recognize signs of cardiac arrest” and “I am able to make available necessary items needed for emergency care whenever on duty.” Readiness was therefore a self-reported measure rather than an objectively observed behavioural outcome (see Supplementary File). The readiness scale demonstrated excellent reliability (α = 0.970). The instrument was pretested at the Winneba Trauma and Specialist Hospital using 50 healthcare professionals comprising 30 nurses and 20 midwives. Participants were recruited from the Emergency Department, medical and surgical wards, theatre/recovery unit, antenatal clinic, labour ward and postnatal ward. A purposive sampling approach was used to recruit nurses and midwives whose clinical roles and work settings were comparable to those of participants in the main study.

Data collection procedures and ethics

Ethical standards were adhered to, and the study was conducted in accordance with the principles of the Declaration of Helsinki. Written informed consent was obtained from all participants. Participants were informed that participation was voluntary and that they could withdraw from the study at any time without penalty. They were also informed that they could decline to answer questionnaire items they were uncomfortable answering without consequence.

Anonymity and confidentiality were maintained through the use of coded identifiers. Hard-copy data were stored in a locked cabinet, while electronic data were stored on a password-protected storage device accessible only to the research team. Ethical approval was obtained from the Hospital Ethical Review Committee (****ERC/EC/2025/005). Recruitment and data collection were conducted from 13 January to 17 February 2025, following institutional clearance and ethical approval. Participants were approached through their respective unit in-charges, and two trained research assistants facilitated recruitment and data collection. After the study objectives and procedures had been explained, written informed consent was obtained before administration of the instruments.

Participants required approximately 20–30 min to complete questions covering demographic characteristics, CPR knowledge, and self-reported readiness to perform CPR. Observed CPR skills were assessed separately through a standardised manikin-based practical demonstration lasting approximately six minutes per participant. Two assessors independently scored participants using the same observational checklist to minimise assessor variation.

Data processing and analysis

Data were checked for completeness and accuracy before analysis. Statistical analyses were conducted using Jamovi version 2.7.12. The missing data were handled by listwise deletion where participant is excluded entirely from the analysis. Descriptive statistics, including frequencies, percentages, means, standard deviations, minimum and maximum values, were used to summarise participants’ socio-demographic characteristics and CPR knowledge, observed skills and self-reported readiness. Chi-square tests of independence were used to examine bivariate associations between participants’ characteristics, CPR knowledge level, observed CPR skill level and self-reported CPR readiness. Statistical significance was assessed at p < .05. Subsequently, binomial logistic regression was conducted to identify factors independently associated with moderate versus low self-reported CPR readiness. The dependent variable was self-reported readiness, with moderate readiness treated as the outcome category and low readiness as the comparison category. The multivariable model included age, sex, educational level, professional category, previous CPR training, CPR knowledge level and observed CPR skill level. These variables represented demographic, professional, training and CPR competency characteristics considered relevant to readiness.

Categorical predictors were entered using specified reference categories. For age, respondents aged ≥ 30 years (older) were the reference group and those aged < 30 years were classified as younger. Male was the reference category for sex, Diploma for educational level, Nurse for professional category, Yes for previous CPR training, low knowledge for CPR knowledge level and low/inadequate skill for CPR skill level. Regression results were reported using the unstandardised logistic coefficient (B), standard error (SE), adjusted odds ratio (AOR), 95% confidence intervals and p-values. Model explanatory performance was assessed using Nagelkerke’s R².

Prior to the multivariable analysis, multicollinearity among the predictors was assessed using variance inflation factor (VIF) and tolerance statistics. VIF values ranged from 1.02 to 1.19 and tolerance values ranged from 0.839 to 0.984, indicating no evidence of problematic multicollinearity. Because the final analysis used binomial logistic regression with categorical predictors, normality of the dependent variable and residuals, homoscedasticity and linear-regression assumptions were not required. The assumption of linearity in the logit for continuous predictors was also not applicable because the predictors included in the final model were categorical. Independence of observations was maintained because each participant contributed one set of responses to the analysis.

Results

Socio-demographic characteristics

Out of the 271 respondents, 224 (82.7%) were females and the rest were males, 47(17.3%). The mean age of respondents was 30.5 years (SD = 1.82). Most respondents were aged 30 years or older (61.6%, n = 167), while 104 respondents were younger than 30 years. Nurses were more (154, 56.8%) than midwives. In terms of educational qualification, majority (51.3%, 139) of the respondents had diploma and those with master’s degree were the minority (0.7%, 2). For work experience, majority (41.7%, 113) have worked for less than 5 years whiles those who have worked for between 16 and 20 years formed the minority (2.6%, 7) (see Table 1).

Table 1.

Demographic characteristics of respondents (N = 271)

Variable	Category	Frequency (N)	Percentage (%)
Sex	Male	47	17.3%
	Female	224	82.7%
Age (years)	< 30 years	104	38.4
	≥ 30 years	167	61.6
Level of Education	Diploma	139	51.3%
	Degree	125	46.1%
	Master’s Degree	2	0.7%
	Others	5	1.8%
Type of Professional	Nurse	154	56.8%
	Midwife	117	43.2%
Working Experience	< 5 years	113	41.7%
	5–10 years	112	41.3%
	11–15 years	39	14.4%
	16–20 years	7	2.6%
CPR Training	Yes	241	88.9%
	No	30	11.1%

Knowledge level

Knowledge level was assessed using 22 CPR knowledge items, with each correct response scored 1 and each incorrect response scored 0, producing a possible total score of 0–22. Total knowledge scores were conceptualised using the modified Bloom classification as poor/low knowledge (< 60%; scores 0–13), moderate knowledge (60%–79%; scores 14–17) and good/high knowledge (≥ 80%; scores 18–22). Descriptive statistics were used to summarise the overall knowledge score, while frequencies and percentages were used to determine the distribution of respondents across the three knowledge categories. Among the 271 nurses and midwives, the mean knowledge score was 10.10 (SD = 1.90), with observed scores ranging from 6 to 14 (see Table 2). The majority of respondents, 257 (94.8%), were classified as having poor/low CPR knowledge, while 14 (5.2%) demonstrated moderate knowledge (see Fig. 1). No respondent attained the good/high knowledge category (see Fig. 1). The findings indicate that CPR knowledge among the respondents was predominantly low based on the modified Bloom classification. (see Table 3)

Table 2.

Level of knowledge of CPR among nurses and midwives

Variable	Category	n	%	M	SD	Minimum	Maximum
Knowledge level	Low (0–13)	257	94.8	10.10	1.90	6.00	14.00
	Moderate (14–17)	14	5.2

Fig. 1.

Level of knowledge of CPR among nurses and midwives

Table 3.

Knowledge of CPR among nurses and midwives

No.	Knowledge question	Correct (n, %)
1	CPR should be done on every patient in a cardiac or respiratory emergency	64 (23.6%)
2	CPR supports and maintains breathing and circulation for an infant, child or adolescent	265 (97.8%)
3	CPR should be conducted on the patient immediately before informing the doctor	19 (7.0%)
4	Ambu bagging is commenced when a patient is gasping or ceases to breathe immediately	35 (12.9%)
5	CPR is initiated when a patient is pulseless or CPR is initiated when a child’s pulse is below 60 bpm	230 (84.9%)
6	Infant cardiac arrest happens more frequently due to respiratory arrest, not due to the heart	225 (83.0%)
7	The reason respiratory arrest leads to cardiac arrest is insufficient oxygen provided to the heart muscle	245 (90.4%)
8	Upon discovering an infant in cardiac arrest, the first step is to seek assistance from other medical teams and assess consciousness	59 (21.8%)
9	I take 25 s to check for pulse of an adult before commencing CPR	123 (45.4%)
10	Check the pulse from the brachial artery within 5 to 10 s	217 (80.1%)
11	The correct position to administer chest compression is immediately below the centre, aligned with the nipples	37 (13.7%)
12	When administering chest compressions, the fingers should be placed vertically to the sternum	21 (7.7%)
13	Chest compressions should be administered at a rate of over 100 times per minute and to a depth of about 4 cm	99 (36.5%)
14	Chest compressions should be 7 inches (10 cm) deep for children during CPR	142 (52.4%)
15	When spinal injuries are suspected, the patient should be placed in a neutral spine position with head fixation	45 (16.6%)
16	The airway of infants should be neutral or head slightly flexed	22 (8.1%)
17	Artificial respiration should be performed twice after assessing the pulse of the infant	62 (22.9%)
18	During artificial respiration, the mouth and nose of the infant should be covered entirely with the face mask	254 (93.7%)
19	Observing the rise and movement of the infant’s chest during artificial breathing is important	260 (95.9%)
20	The ratio of chest compressions to artificial breathing should be 30:2 in infants	34 (12.5%)
21	During adult CPR, four cycles of 30 chest compressions and 2 ventilations should be completed within one minute.	30 (11.1%)
22	CPR (BSL) can be ceased when the infant breathes independently when advanced help arrives	262 (96.7%)

Level of CPR skills

CPR skill level was assessed using a 12-item observational checklist, with each correctly demonstrated skill scored 1 and each incorrectly performed or omitted skill scored 0, producing a possible total score of 0–12 (see Table 4). Two trained research assistants independently observed and scored participants using the same checklist and predefined scoring criteria. Where discrepancies occurred between the two assessors, the specific skill item was reviewed against the established scoring criteria and discussed by both assessors until consensus was reached. The agreed score was then recorded as the participant’s final score for that item. For the purpose of classification, respondents were categorised using a 60% threshold, whereby scores of 0–7 were classified as inadequate CPR skill and scores of 8–12 as adequate CPR skill. Descriptive statistics were used to summarise the overall skill score, while frequencies and percentages were used to describe respondents’ performance on individual CPR skill items and their overall skill categories (see Table 5).

Table 4.

Observed CPR skills among nurses and Midwives (N = 271)

No.	CPR skill item	Correct, n (%)
1	Ensure safety of the environment	151 (55.7)
2	Confirm that the patient is not breathing and/or is pulseless	225 (83.0)
3	Call for help	185 (68.3)
4	Assess for trauma and move the patient to a firm surface	134 (49.4)
5	Tilt the patient’s head into an appropriate position to keep the airway open	217 (80.1)
6	Position appropriately and administer rescue breaths effectively	203 (74.9)
7	Determine the appropriate site for heart massage/chest compression	179 (66.1)
8	Use the appropriate compression technique by placing two fingers perpendicular to the chest or wrapping both hands around the chest with both thumbs positioned midline below the nipples	175 (64.6)
9	Apply pressure to achieve the appropriate compression depth and compression-to-ventilation ratio according to the patient’s age	184 (67.9)
10	Apply the previous steps to achieve 100–120 compressions per minute	167 (61.6)
11	Observe for chest expansion when performing bag-and-mask ventilation	208 (76.8)
12	Continue basic life support until the patient breathes independently or advanced medical help arrives	241 (88.9)

Table 5.

Level of skill of CPR among nurses and midwives

Variable	Category	n	%	M	SD	Minimum	Maximum
CPR skill level	Low	93 (0–7)	34.3	8.37	3.25	1.00	12.00
	High	178 (≥ 8)	65.7

Among the 271 nurses and midwives, the mean CPR skill score was 8.37 (SD = 3.25), with observed scores ranging from 1 to 12 (see Table 5). Overall, 178 respondents (65.7%) were classified as having high/adequate CPR skills, whereas 93 (34.3%) had low/inadequate skills (see Fig. 2). At the item level, the highest proportions of correct performance were observed for continuing basic life support until the patient breathed independently or advanced help arrived (88.9%, n = 241), confirming that the patient was not breathing and/or was pulseless (83.0%, n = 225), and maintaining an open airway through appropriate head positioning (80.1%, n = 217). In contrast, the lowest correct performance was recorded for assessing trauma and moving the patient to a firm surface (49.4%, n = 134), followed by ensuring environmental safety (55.7%, n = 151) and achieving the required compression rate of 100–120 compressions per minute (61.6%, n = 167). The findings indicate that although approximately two-thirds of respondents met the study threshold for adequate CPR skills, there were important gaps in several essential components of CPR performance.

Fig. 2.

Level of skill of CPR among nurses and midwives

Self-reported readiness for CPR practice

Self-reported readiness for CPR practice was assessed using 11 positively worded items rated on a five-point Likert scale from 1 (strongly disagree) to 5 (strongly agree), producing a possible total score of 11–55, with higher scores indicating greater perceived readiness to perform CPR (see Table 6). The construct reflected respondents’ perceived confidence, willingness to act, perceived capability, preparedness, and beliefs about performing CPR rather than objectively observed clinical readiness. Descriptive statistics were used to summarise the overall readiness score and individual item responses, while frequencies and percentages were used to describe respondents according to the readiness categories applied in the study. Among the 271 nurses and midwives, the mean self-reported readiness score was 44.60 (SD = 11.30), with observed scores ranging from 11 to 55 (see Table 7). Overall, 244 respondents (90.0%) were classified as having moderate self-reported readiness, while 27 (10.0%) were classified as having low readiness. At the item level, the highest-rated statements were the belief that prompt initiation of CPR can significantly improve survival (M = 4.42, SD = 1.20), the belief that one’s actions during CPR can make a difference in saving patients’ lives (M = 4.29, SD = 1.14), and readiness to respond to a call for help (M = 4.18, SD = 1.14). The comparatively lowest ratings were for knowledge of the correct compression-to-ventilation ratio (M = 3.86, SD = 1.18), perceived training in providing adequate ventilation and high-quality chest compressions (M = 3.89, SD = 1.12), and confidence in coordinating CPR efforts within a team (M = 3.92, SD = 1.13). Overall, the findings indicate that respondents generally reported positive perceptions of their readiness to perform CPR, although some areas of perceived technical preparedness remained comparatively weaker.

Table 6.

Readiness for CPR practice (N = 271)

Items on Readiness	Mean ± SD
I believe that prompt initiation of CPR can significantly improve survival rates	4.42 ± 1.20
I am confident in my ability to recognize signs of cardiac arrest	4.08 ± 1.11
I am able to make available necessary items needed for emergency care whenever on duty	3.93 ± 1.18
I am always ready to respond to a ‘call for help’ whenever the need be	4.18 ± 1.14
I am trained to provide adequate ventilation and to give high quality chest compressions	3.89 ± 1.12
I am knowledgeable about the correct compression to ventilation ratio for CPR	3.86 ± 1.18
I am ready to begin CPR before calling doctors to assist	4.06 ± 1.18
I am committed to staying up-to-date with CPR guidelines and best practices through regular training	4.08 ± 1.18
I believe that my actions in performing CPR can make difference in saving patients’ lives	4.29 ± 1.14
I am confident in my ability to effectively communicate and coordinate CPR efforts within a team	3.92 ± 1.13
I am willing to perform CPR on all patients including neonates, infants, children and adolescents	3.95 ± 1.15

Table 7.

Level of self-reported Readiness of CPR among nurses and midwives

Variable	Category	n	%	M	SD	Minimum	Maximum
Self-reported readiness level	Low	27	10.0	44.60	11.30	11.00	55.00
	Moderate	244	90.0

Association between age, knowledge, skills and self-reported readiness to practice CPR

Chi-square test of independence was conducted to examine associations between participant characteristics, CPR knowledge, CPR skills and self-reported CPR readiness among 271 nurses and midwives (see Table 8). CPR knowledge level was significantly associated with readiness, χ²(1, N = 271) = 36.60, p < .001; 92.6% of respondents with low knowledge had moderate readiness compared with 42.9% of those with moderate knowledge. CPR skill level was also significantly associated with readiness, χ²(1, N = 271) = 34.40, p < .001; 97.8% of respondents with high CPR skills had moderate readiness compared with 75.3% of those with low skills. However, no significant associations were found between self-reported readiness and sex, χ²(1) = 3.16, p = .076; educational level, χ²(2) = 2.01, p = .367; professional category, χ²(1) = 1.18, p = .277; working experience, χ²(3) = 4.21, p = .240; or previous CPR training, χ²(1) < 0.01, p = .994. Overall, CPR knowledge and CPR skill level were the only variables significantly associated with self-reported CPR readiness (see Table 8).

Table 8.

Association between participant characteristics, CPR knowledge, CPR skills, and self-reported CPR readiness (N = 271)

Variable	Category	Low readiness, n (%)	Moderate readiness, n (%)	χ²	p
Sex	Male	8 (17.0)	39 (83.0)	3.16	0.076
	Female	19 (8.5)	205 (91.5)
Educational level	Diploma	17 (12.2)	122 (87.8)	2.01	0.367
	Degree	9 (7.2)	116 (92.8)
	Master’s degree	1 (14.3)	6 (85.7)
Professional category	Nurse	18 (11.7)	136 (88.3)	1.18	0.277
	Midwife	9 (7.7)	108 (92.3)
Working experience	Less than 5 years	16 (14.2)	97 (85.8)	4.21	0.240
	5–10 years	8 (7.1)	104 (92.9)
	11–15 years	3 (7.7)	36 (92.3)
	16–20 years	0 (0.0)	7 (100.0)
Previous CPR training	Yes	24 (10.0)	217 (90.0)	< 0.01	0.994
	No	3 (10.0)	27 (90.0)
CPR knowledge level	Low	19 (7.4)	238 (92.6)	36.60	< 0.001
	Moderate	8 (57.1)	6 (42.9)
CPR skill level	Low	23 (24.7)	70 (75.3)	34.40	< 0.001
	High	4 (2.2)	174 (97.8)

Afterwards, a binomial logistic regression was conducted to identify factors associated with moderate versus low self-reported CPR readiness among 271 nurses and midwives. The model accounted for approximately 35.0% of the variation in readiness according to Nagelkerke’s R² = 0.350. After adjustment for age, sex, educational level, professional category, previous CPR training, CPR knowledge level, and CPR skill level, only CPR knowledge and CPR skill level were statistically significant predictors. Respondents with moderate CPR knowledge had significantly lower log odds of moderate rather than low readiness compared with those with low knowledge, B = -2.24, SE = 0.75, 95% CI [-3.70, -0.78], p = .003, corresponding to an AOR of 0.11, 95% CI [0.03, 0.46]. In contrast, respondents with high CPR skill had significantly higher odds of moderate rather than low readiness compared with those with low skill, B = 2.55, SE = 0.58, 95% CI [1.41, 3.70], p < .001, AOR = 12.84, 95% CI [4.09, 40.36]. Age, sex, educational level, professional category, and previous CPR training were not significantly associated with self-reported CPR readiness (all p > .05) (see Table 9).

Table 9.

Binary logistic regression of Predictors on self-reported Readiness

Variable	Category	B	95% CI for B LL	95% CI for B UL	SE	AOR	95% CI for AOR LL	95% CI for AOR UL	p
Model fit	Nagelkerke R² = 0.350
Age	Older	Ref	Ref	Ref	Ref	Ref	Ref	Ref	Ref
	Younger	-0.191	-1.164	0.782	0.496	0.826	0.312	2.185	0.700
Sex	Male	Ref	Ref	Ref	Ref	Ref	Ref	Ref	Ref
	Female	0.165	-1.188	1.518	0.690	1.180	0.305	4.562	0.811
Educational level	Diploma	Ref	Ref	Ref	Ref	Ref	Ref	Ref	Ref
	Degree	0.701	-0.299	1.700	0.510	2.015	0.742	5.476	0.169
	Master’s degree	0.788	-2.002	3.579	1.424	2.200	0.135	35.839	0.580
Professional category	Nurse	Ref	Ref	Ref	Ref	Ref	Ref	Ref	Ref
	Midwife	0.145	-0.922	1.212	0.544	1.156	0.398	3.360	0.790
Previous CPR training	Yes	Ref	Ref	Ref	Ref	Ref	Ref	Ref	Ref
	No	0.674	-0.805	2.153	0.755	1.962	0.447	8.614	0.372
CPR knowledge level	Low	Ref	Ref	Ref	Ref	Ref	Ref	Ref	Ref
	Moderate	-2.236	-3.696	-0.776	0.745	0.107	0.025	0.460	0.003
CPR skill level	Low	Ref	Ref	Ref	Ref	Ref	Ref	Ref	Ref
	High	2.553	1.408	3.698	0.584	12.844	4.088	40.360	< 0.001

Discussion

Summary of findings

The findings are discussed in relation to the study objectives and compared with relevant studies globally, in African and particularly in Ghana. The discussion focuses on CPR knowledge, observed skills, self-reported readiness and factors associated with readiness, while also considering measurement limitations that may have influenced the findings.

The principal findings showed that CPR knowledge was generally poor, whereas observed practical skills were comparatively better and self-reported readiness was generally favourable. At the bivariate level, CPR knowledge and observed skills were significantly associated with readiness. In the adjusted analysis, observed CPR skill showed the strongest relationship with readiness, whereas the relationship involving knowledge was less straightforward because almost all participants were classified as having low knowledge. Age and the other demographic and training-related characteristics examined were not significantly associated with self-reported readiness after adjustment.

CPR knowledge

The findings indicate a deficit in CPR knowledge among the nurses and midwives. Overall, 94.8% had low knowledge, 5.2% had moderate knowledge and none attained the high-knowledge category. Although respondents performed well on some general concepts, there were gaps in the sequence of resuscitation, compression position and technique, compression rate and depth, infant airway positioning, compression-to-ventilation ratios, and paediatric CPR. These are clinically important deficits because current guidelines emphasise rapid recognition and emergency activation, immediate high-quality compressions, appropriate ventilation, correct compression depth and rate, minimal interruptions, and age-appropriate resuscitation procedures [3, 17–20].

Globally, the low knowledge level is consistent with evidence that CPR knowledge and skills decline after For instance, nurses in Botswana had a baseline knowledge score of approximately 48%, closely resembling the 45.9% obtained in this study and improvements after training diminished by six months [14]. Poor or inadequate CPR knowledge has similarly been reported in Tanzania, Namibia, South Africa and Ethiopia [21–25]. In southwest Nigeria, the mean knowledge score was 43.6%, with particularly poor understanding of CPR sequence and compression-to-ventilation ratios [26]. Importantly, intervention studies demonstrate that these deficits are modifiable as structured training produced considerable improvements among nurses in Uganda, Nigeria and Mozambique, although deterioration after training remained a concern [27–29]. These findings imply that the problem is better understood as one of competency acquisition and retention rather than an immutable professional deficiency.

Evidence in Ghana is mixed but still useful. Nti-Darkwah et al. [12] found that only 39.9% of nurses demonstrated good advanced life-support knowledge, while other studies have reported higher CPR knowledge, particularly among emergency-department personnel or participants assessed soon after training [30–32]. For example, 74.6% of participants in the Ghana Heart Initiative achieved adequate BLS knowledge and skills immediately following structured training [31], while Quao et al. [33] also demonstrated improved BLS scores following training at Korle Bu Teaching Hospital. Earlier Ghanaian evidence likewise identified CPR knowledge gaps and the need for periodic training [34]. The contrasting findings may also be due to differences in clinical setting, adult versus paediatric CPR content, measurement instruments, scoring thresholds and participants’ CPR exposure. However, because the recency and frequency of previous CPR training were not adequately characterised in this study, differences in training recency should not be used to explain the observed knowledge level.

Moreover, the restricted distribution of knowledge scores is also important when interpreting associations with self-reported readiness. With 94.8% of respondents classified as having low knowledge, only a small proportion remained in the moderate category. This limited variability reduced the ability of the analysis to distinguish how progressively higher levels of CPR knowledge might relate to readiness and should be considered when interpreting the apparently counterintuitive adjusted association between knowledge and readiness.

CPR skills

Observed CPR skills were considerably better than theoretical knowledge. The respondents performed relatively well in continuing BLS until spontaneous breathing or advanced assistance arrived, recognising absent breathing or pulse, maintaining an open airway and observing chest expansion. However, important deficiencies remained. For instance, only 49.4% correctly assessed trauma and moved the patient to a firm surface, 55.7% ensured environmental safety, 61.6% achieved the required compression rate, and only about two-thirds demonstrated the correct compression site and technique. Therefore, meeting the 60% threshold should not be interpreted as complete guideline concordant CPR competence. This pattern is plausible because cognitive knowledge, psychomotor performance and self-confidence are related but distinct dimensions of competence, and theoretical knowledge does not always correspond with objectively demonstrated CPR performance [35, 36].

The 65.7% adequate skill level compares favourably with some African studies reporting lower practical competency among healthcare workers in Botswana, Tanzania, Ethiopia and South Africa, although differences in checklists, pass marks and simulated scenarios limit direct comparison [14, 21, 23, 24]. Evidence in Ghana also shows that CPR performance is influenced by confidence, timely initiation, availability of equipment and medications, workplace ergonomics and institutional procedures [13], while structured practical training and greater resuscitation exposure may improve performance [30, 31, 33]. The findings therefore imply that training should target the specific psychomotor deficits identified through frequent manikin-based practice, deliberate practice, feedback and periodic direct reassessment of compression rate and depth, technique, airway positioning, ventilation, patient positioning, environmental safety and team response [37–40].

The classification of 65.7% of respondents as having adequate skills should nevertheless be interpreted cautiously because adequacy was defined using the 60% threshold set by this study. This cut-off meant that a participant could fail several important steps yet still be classified as having adequate overall skill. Consequently, the dichotomised skill variable may have combined borderline performers with participants demonstrating more consistently competent performance, reducing discrimination between levels of actual CPR competence.

Self-reported readiness to practise CPR

The respondents generally showed moderate self-reported readiness to perform CPR. They believed that CPR could improve survival and were willing to respond when needed. However, they showed relatively lower confidence in some technical aspects of CPR and in coordinating effectively with a team. Overall, willingness to perform CPR was stronger than confidence in technical and team-based performance. This difference is consistent with global and African evidence demonstrating that readiness and knowledge are distinct and not interchangeable. For instance, Saudi Arabian nurses showed high self-efficacy despite only moderate knowledge [41]. Similarly, favourable attitudes have coexisted with weaker CPR performance in Namibia, Botswana and South Africa [14, 22, 25]. Riggs et al. [36] also demonstrated limited correspondence between self-efficacy and assessed CPR skill. This supports the conclusion that readiness should be interpreted as perceived capability and willingness, not proof of actual technical competence.

Evidence in Ghana supports this interpretation as Amoako-Mensah et al. [13] identified confidence, knowledge, practical skills, timely initiation, equipment availability, team functioning and environmental conditions as important influences on CPR quality, while Anto-Ocrah et al. [42] similarly found that knowledge, training and perceived ability influenced willingness to perform CPR. Although readiness in this study appears more favourable than the 39.8% confidence reported by Nti-Darkwah et al. [12], the measures and clinical focus differed, while Sedem et al. [32] also reported generally favourable attitudes among Ghanaian nurses. The findings therefore suggest that respondents’ willingness to perform CPR provides a useful foundation for improvement, but perceived readiness should not be regarded as evidence of technical competence. Simulation, objective performance assessment, feedback and debriefing are needed to ensure that confidence is matched by demonstrated CPR ability.

This distinction is important for interpretation of the present findings. The skill assessment reflected directly observed performance on a structured checklist, whereas readiness reflected participants’ own perceptions of their willingness and ability to act. A respondent may therefore report being ready to initiate CPR while still performing some technical components incorrectly or may demonstrate the required skills while expressing uncertainty about performing them during an actual emergency. The two outcomes should consequently be treated as complementary but non-equivalent dimensions of CPR preparedness.

Factors associated with CPR readiness

Both knowledge and skill were significantly associated with readiness at the bivariate level. However, observed CPR skill showed the strongest adjusted relationship. Participants with adequate skills had substantially higher odds of moderate readiness than those with inadequate skills, suggesting that practical competence is closely linked with perceived readiness to respond. This is consistent with evidence that hands-on practice, simulation, mastery and feedback can improve both CPR performance and confidence [38–40, 43]. Intervention studies have similarly reported improvements in CPR knowledge and skills following structured training [14, 27, 28], while evidence in Ghana indicates that skill and confidence are important components of resuscitation competence [13]. This finding therefore supports emphasis on practical, simulation-based CPR training rather than relying mainly on theoretical instruction.

Nevertheless, the association should not be interpreted as showing that skill and readiness are similar. This is because better performance may support confidence and perceived ability, but the cross-sectional design cannot establish whether stronger skills increased readiness, whether more confident respondents engaged more successfully with the skill assessment, or whether both reflected other unmeasured influences. Moreover, the 60% skill threshold may have weakened the precision of this relationship by grouping participants with substantially different levels of performance within the same “adequate” category.

Moreover, the adjusted relationship between knowledge and readiness was unexpected, as participants with moderate knowledge had lower odds of moderate readiness than those with low knowledge. This should not be interpreted as evidence that greater knowledge reduces readiness because the moderate knowledge group was very small, making the estimate vulnerable to sparse-data instability. The finding also differs from studies that have linked better knowledge with improved self-efficacy or practice [25, 41]. In the present study, the unequal distribution of knowledge categories also produced considerable restriction of range, with almost all respondents classified as having low knowledge. This reduces confidence that the categorical analysis captured a genuine relationship between knowledge and readiness. In addition, knowledge of CPR procedures and perceived readiness to intervene represent different components of preparedness; therefore, a strong relationship should not automatically be expected. A possible explanation is that more knowledgeable respondents may assess their own preparedness more critically, but this was not directly examined in the study. Accordingly, this explanation is speculative and the finding is better interpreted in light of the small moderate-knowledge group and limitations of the knowledge classification.

In addition, sex, educational level, professional category, work experience, previous CPR training and age were not significantly associated with self-reported readiness after adjustment. Similar nonsignificant relationships for some demographic characteristics have been reported in Nigeria and Botswana [14, 26], although other studies have associated education, recent training and experience with better CPR knowledge or competence [12, 21, 24].

The absence of a significant association between age and readiness suggests that chronological age alone was not an important independent determinant of perceived CPR preparedness in this sample. Age may reveal accumulated professional or clinical exposure, but it does not directly measure CPR-specific experience, current competence or confidence. The finding therefore should not be interpreted as showing that age can never influence CPR performance but rather, after accounting for the variables included in this study, age did not independently explain differences in self-reported readiness.

Importantly, the lack of association between previous CPR training and readiness does not indicate that training is ineffective because a simple history of whether training had ever occurred does not establish when training occurred or provide sufficient information about the content, duration or quality of that training. Therefore, the present analysis can only address whether having a history of CPR training was associated with readiness but it cannot determine whether training recency, frequency or intensity influenced readiness. However, studies have recorded that, improvements declined after training and deteriorated over time [14, 31, 44]. Therefore, training records should capture when training occurred, its format and frequency and whether competence was reassessed.

These findings demonstrate a knowledge–skill–readiness gap where theoretical CPR knowledge was generally poor, practical skills were better but incomplete, while self-reported readiness was comparatively favourable. Importantly, the three findings should not be interpreted as directly interchangeable indicators of competence because observed skill represented demonstrated psychomotor performance whereas readiness represented perceived preparedness. The findings further suggest that CPR preparedness should be approached as an integrated competency involving knowledge, practical performance, confidence, teamwork and an enabling clinical environment. There should also be scenario-based simulation, spaced practice, objective feedback, team drills and debriefing to address weaknesses in compression rate and technique, airway and patient positioning, ventilation, scene safety and team coordination [17, 18, 43, 45, 46]. Competency assessment should also combine written testing with directly observed performance, supported by visible algorithms, functional bag-mask devices, accessible resuscitation equipment and clear emergency escalation procedures [13].

Limitations of the study

This study has several limitations that should be considered when interpreting the findings. First, the cross-sectional design means that the observed relationships between CPR knowledge, observed skills and self-reported readiness cannot establish causality or determine the direction of the associations. Second, readiness was self-reported and may have been influenced by social desirability or inaccurate self-assessment, and therefore should not be interpreted as equivalent to objectively demonstrated CPR competence. Third, the knowledge measure showed relatively low internal consistency (KR-20 = 0.475), while the moderate-knowledge group was very small compared with the low-knowledge group. These measurement issues may have reduced the stability of the estimated association between knowledge and readiness. Fourth, although CPR skills were directly observed, performance on a manikin may not fully reflect performance during real clinical emergencies. In addition, the 60% threshold for adequate skill was relatively permissive and may have classified some borderline performers as adequate, thereby reducing discrimination between different levels of competence. Previous CPR training was measured as whether participants had ever received training and did not adequately capture training recency, duration, frequency, quality or simulation content, therefore, the effects of these training characteristics could not be examined. Finally, because the study was conducted among nurses and midwives in a single tertiary hospital, caution is required when generalising the findings to other hospitals, primary healthcare facilities or healthcare professionals in Ghana and other settings where staffing, emergency resources, training opportunities and exposure to cardiac arrest may differ.

Conclusion

Cardiopulmonary resuscitation preparedness among nurses and midwives is characterised by a knowledge–skill–readiness gap. Although the majority of respondents demonstrated low CPR knowledge, approximately two-thirds achieved the adequate level of observed skills, while most reported moderate readiness to perform CPR. This indicates that knowledge, practical competence and perceived preparedness are related but distinct dimensions and should not be used interchangeably when assessing CPR competence. The particularly strong positive association between observed skill and readiness suggests that practical mastery may be central to developing confidence and preparedness, whereas the unexpected inverse association between moderate knowledge and readiness should be interpreted cautiously because of the small number of respondents in the moderate-knowledge category. Consequently, effective preparedness requires more than previous certification or willingness to intervene but require up-to-date knowledge, demonstrated psychomotor competence, effective teamwork and a healthcare environment that supports rapid and high-quality resuscitation.

Recommendations for policy, practice and further research

The Ministry of Health, Ghana Health Service, Teaching Hospitals, Nursing and Midwifery Council of Ghana and other relevant regulatory bodies should strengthen requirements for maintaining CPR competence among nurses and midwives. Periodic competency renewal should combine knowledge assessment with directly observed practical performance, while refresher CPR training should be integrated into continuing professional development. Healthcare facilities should also maintain current CPR algorithms, functional emergency equipment, regular resuscitation drills and detailed records of staff training and competency.

In addition, brief and recurrent departmental CPR simulation should be introduced, with emphasis on the specific weaknesses identified in the study, including compression rate and technique, airway and patient positioning, ventilation, environmental safety, compression-to-ventilation ratios and team coordination. Regular team drills, structured debriefing, manikin-based practice, accessible resuscitation equipment and clearly displayed CPR algorithms should be used to strengthen practical competence and translate willingness to perform CPR into effective performance.

Further research should use multicentre, longitudinal and intervention designs across different levels of healthcare in Ghana. Future studies should assess the recency, frequency, duration and type of CPR training, simulation exposure, certification and actual cardiac-arrest experience, while also examining continuous knowledge, skill and readiness scores. Research should further explore real clinical CPR performance and organisational factors such as workload, staffing, equipment availability, leadership and communication and compare refresher-training approaches to identify effective and sustainable models for Ghanaian healthcare settings.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

The study was conceived by RO-D and MAB. RO-D, MAB and EQ contributed to the design of the study. RO-D, MAB, GO and JB contributed to the acquisition of data. RO-D, EO, JB, GO and EQ contributed to the data analysis and interpretation. All the authors wrote the manuscript and revised it for important intellectual content. They all read and approved the final version for submission to this journal.

Funding

The study was solely funded by the authors.

Data availability

The datasets used and/or analysed during the current study is available in the OSF Data repository DOI:10.17605/OSF.IO/HMN9Z.

Declarations

Ethics approval and consent to participate

Ethical standards were strictly adhered to, and the study was conducted in accordance with the principles of the Declaration of Helsinki. Ethical approval was obtained from the Hospital Ethical Review Committee (ID: ****ERC/EC/2025/005). Written informed consent was obtained from all participants. Participants were assured of their voluntary participation and right to withdraw at any time without penalty. They were assured of confidentiality and anonymity and confidentiality were maintained using coded identifiers.

Consent for publication

This is not applicable.

Competing interests

The authors declare no competing interests.