Development and validation of the knowledge, attitudes, and practices (KAP) questionnaire on pediatric pain management for physicians and nurses

Abstract

Background

Effective pediatric pain management is often hindered by gaps in medical staff’s knowledge, attitudes, and practices. Few reliable instruments comprehensively measure these domains. This study aimed to develop and preliminarily validate a Knowledge, Attitudes, and Practices (KAP) Questionnaire on pediatric pain management for physicians and nurses.

Methods

A preliminary pool of items was generated from a literature review and refined through two rounds of Delphi consultations with 19 multidisciplinary experts. A pilot test with 20 physicians and nurses was conducted to ensure clarity. In June 2024, a convenience sample of 603 physicians and nurses from a specialized children’s hospital in Chongqing, China, completed the questionnaire for psychometric validation.

Results

The final questionnaire comprised 32 items across three dimensions (knowledge: 13; attitudes: 10; practices: 9), a self-report tool using a five-point Likert scale. Items showed good discrimination (CR: 12.29–19.78) and acceptable item–total correlations (r = 0.49–0.72). Content validity indices were high (I-CVI: 0.86–1.00; S-CVI/Ave: 0.996). The overall Cronbach’s α was 0.96, with values for the three dimensions ranging from 0.92 to 0.98. Exploratory factor analysis extracted three factors explaining 70.56% of the variance (KMO = 0.95; Bartlett’s χ² = 12,459.51, p < 0.001), supporting construct validity, and confirmatory factor analysis supported the three-factor structure (CFI = 0.90, TLI = 0.90, RMSEA = 0.10, SRMR = 0.06). Convergent validity was supported by average variance extracted values ranging from 0.56 to 0.84. Discriminant validity was established as the square root of the AVE for each factor (knowledge: 0.79; attitudes: 0.92; practices: 0.75) exceeded its inter-factor correlations.

Conclusions

The pediatric pain management KAP questionnaire demonstrated adequate psychometric properties in this initial validation. It provides a practical tool to assess physicians’ and nurses’ knowledge, attitudes, and practices, supporting the identification of gaps and guiding targeted training and quality improvement initiatives to enhance pediatric pain care.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12913-026-14932-1.

Introduction

Pain is one of the most common and distressing symptoms experienced by hospitalized children. Untreated pain not only causes acute changes in children’s physiological and behavioral states and hormone levels [1], potentially worsening their condition and prolonging hospitalization, but it may also exacerbate anxiety and fear, undermining the trust relationship among children, families and healthcare providers [2]. Furthermore, repeated or persistent painful stimuli can have long-term implications, leading to the development of chronic and complex pain problems, as well as medical phobias and avoidance behaviors [3]. The evidence is clear that pain management should be considered just as crucial as any other part of medical procedures and treatments. The International Association for the Study of Pain (IASP) has launched a global initiative called ChildKind, aimed at encouraging healthcare professionals in medical institutions worldwide to provide comfort and pain relief for children [4].

Despite the availability of evidence-based guidelines, pain in pediatric patients remains under-recognized and undertreated [5]. A national survey conducted in China across 512 pediatric units (including NICU, PICU, general inpatient wards, and outpatient clinics) reported that 60.98% of these units did not routinely conduct pain assessments [6]. Internationally, inadequate analgesia has also been documented: a study at a tertiary hospital in Ethiopia found that 63.30% of hospitalized children received insufficient pain medication, with the type of analgesic prescribed often mismatched to the child’s reported pain severity [7].

A major challenge in pediatric pain management, as in pediatric healthcare more broadly, is how to effectively translate best practices into routine care. The key to addressing this issue is changing physicians’ and nurses’ pain management behaviors based on evidence. The Knowledge-Attitude-Practice (KAP) model, first introduced by Professor Mayo at Harvard University, explains that behavior change is a sequential process involving knowledge acquisition, belief formation, and behavior modification [8]. Interventions based on the KAP model can enhance the effectiveness of pain management and improve patient outcomes [9]. Therefore, understanding and then improving physicians’ and nurses’ KAP regarding pain management is essential for closing the gaps in current pediatric pain management practices. A validated KAP questionnaire can serve this purpose by: (a) identifying modifiable knowledge deficits (e.g., misconceptions about opioid safety in children), (b) detecting attitude barriers (e.g., beliefs that infants do not feel pain), and (c) revealing gaps between knowledge and clinical practice, thereby informing targeted training and quality improvement strategies.

Despite the availability of multiple instruments, there remains no reliable and standardized KAP measure for pediatric pain management applicable across both physicians and nurses in general clinical settings. The most widely used tool, the Pediatric Nurses’ Knowledge and Attitudes Survey (PNKAS), was designed specifically for nurses and measures only knowledge and attitudes, leaving practices unexamined [10]. Its content also predates contemporary guidance and does not fully reflect current priorities, including family participation in pain care [11]. Other existing questionnaires are tailored to specific clinical scenarios, such as neonatal intensive care [12], community practice [13], and perioperative care [14]. While valuable within their intended contexts, these tools have limited applicability beyond those settings, making cross-context comparisons difficult. In addition, many existing instruments include numerous items and complex formats, increasing respondent burden and lowering response rates [15], which undermines their utility for routine clinical audit or large-scale surveys. Thus, no single existing instrument simultaneously addresses the following needs: (a) unified coverage of knowledge, attitudes, and practices within one questionnaire; (b) applicability to both physicians and nurses without adaptation; (c) brevity suitable for routine clinical use; and (d) relevance across diverse pediatric care settings.

To address these gaps, this study aims to develop and validate a new KAP questionnaire for use by both physicians and nurses, with the same set of items applicable to both professions without role-specific modifications (a core module approach). The instrument is intended to be concise while covering all three KAP domains. A validated KAP questionnaire can serve as a standardized tool for supporting consistent evaluation across different clinical settings and professional groups, providing evidence to guide targeted improvements in pediatric pain management.

Methods

Study design

The development and validation study were carried out in three main steps. First, taking the KAP model as a framework, a literature review was performed to compile a pool of items, followed by the development of the initial questionnaire based on the Delphi method [16]. Second, a pilot test and content validation were conducted to refine the items. Finally, the third step involved a psychometric evaluation of validity and reliability using a sample of pediatric physicians and nurses.

Instrument

Design and creation of the preliminary questionnaire

Using the KAP model as a framework, an initial item pool was developed through a literature review. The research team searched CNKI, Wanfang Data, PubMed, and Embase using terms including “pediatric/child”, “pain”, “assessment/management/practice”, and “knowledge/attitude/behavior”. The search aimed to identify relevant guidelines, expert consensus, KAP studies, and existing instruments to inform item development. Based on the literature findings and clinical practice contexts, the research team preliminatively developed 30 items (10 knowledge, 9 attitude, 11 practice) as the initial item pool. Knowledge and attitude items used a 5-point Likert scale (1 = strongly disagree to 5 = strongly agree), while practice items used a 5-point frequency scale (1 = never to 5 = always).

Subsequently, two rounds of expert consultations were conducted via email or WeChat with experts in the relevant field. The inclusion criteria for experts were as follows: (a) a bachelor’s degree or higher, (b) an associate senior professional title or above, (c) engagement in medical, nursing, management, or research work related to pain management, (d) more than five years of work experience, and (e) willingness to participate in the consultations. In total, an expert panel of 19 professionals from nine provinces and municipalities in China (Beijing, Guangzhou, Shanghai, Zhejiang, Jiangsu, Hubei, Chongqing, Shanxi, and Liaoning) participated in this study. The panel included experts in pediatric nursing, pediatrics, pain management nursing, and pain medicine. The demographic characteristics of the experts are presented in Table 1.

Table 1.

Demographic characteristics of the consulting experts(N = 19)

Variables	n	%	Mean (SD)
Age (years)			50.37(7.58)
Gender
Male	4	21.05
Female	15	78.95
Education level
Doctor	5	26.32
Master	10	52.63
Bachelor	4	21.05
Title
Senior	12	63.16
Deputy Senior	7	36.84
Whether or not a graduate supervisor
PhD advisor	4	21.05
Master’ s advisor	9	47.37
No	6
Work location
Children’s hospital	13	68.42
General Hospital	5	26.32
Maternity and Child Health Hospital	1	5.26
Areas of expertise
Pediatric nursing	7	36.84
Pediatrics	2	10.53
Pain management nursing	7	36.84
Pain medicine	3	15.79
Work seniority (years)			28.58(8.75)

SD = Standard Deviation

The consultation questionnaire had three parts. The first part introduced the purpose, content, and significance of the study to the experts. In the second part, experts were invited to rate the importance of each item using a 5-point Likert scale (5 for very important, 1 for very unimportant) and to provide suggestions for adding, deleting, or modifying items. The third part collected general demographic data of the experts, as well as their self-evaluation of judgment criteria, and familiarity.

The first round of expert questionnaires was collected from May 6 to May 13, 2024. Based on the experts’ suggestions and predetermined rules, the items were revised. Subsequently, the second round of questionnaires was then prepared and distributed to the experts, with responses received from May 22 to May 29, 2024. Items were then supplemented, deleted, or modified based on expert opinions until a consensus was reached. The criteria for item selection in this study were as follows: a mean importance score of no less than 3.50, a coefficient of variation of no more than 0.25, and a full-score ratio of greater than 50%. Items that did not meet all three thresholds were excluded. For those failing one or two thresholds, decisions were made based on expert opinions and group discussions.

All invited experts completed both rounds, yielding a 100% response rate. The coefficient of judgment basis (Ca) and familiarity coefficient (Cs) were 0.95 and 0.93 in the first round, and 0.92 and 0.89 in the second round, respectively, all exceeding 0.75. These results suggest that the panel possessed a high level of expertise in this field, thereby ensuring the credibility of the consultation outcomes. The degree of concentration and coordination of expert opinions is presented in Table 2. Based on expert feedback and predefined criteria, 1 item was deleted, 8 new items were added, and 20 items were revised in the first round. In the second round, 27 items were revised and 1 item was deleted. After two rounds, the coefficient of variation for each item was < 0.25, and the P-values for Kendall’s W test were all < 0.05, indicating good consistency among expert opinions. The finalized version of the initial questionnaire comprised 36 items, including 13 items on knowledge, 10 items on attitudes, and 13 items on practices. This questionnaire is intended for use in the subsequent pilot survey.

Table 2.

Degree of concentration and coordination of expert opinions

Rounds	Mean score	full-score frequency (%)	coefficient of variation	Kendall’s W	Chi-square	P
Round 1	4.53–4.95	52.63% -94.74%	0.05–0.13	0.09	46.87	<0.05
Round 2	4.21–4.95	36.84%-94.74%	0.05–0.17	0.18	123.25	<0.01

Content validity and pilot study

The pilot study was conducted in June 2024 at a tertiary-level pediatric hospital in Chongqing, China. A total of 20 physicians and nurses (10 from each profession), each with over one year of pediatric experience, were recruited through convenience sampling. They completed the initial questionnaire and provided feedback on its readability and comprehensibility. The feedback yielded 26 comments, and the average completion time was six minutes. Based on this feedback, the study team revised the wording of seven items.

Subsequently, content validity was assessed by six experts who had participated in the initial consultation, as they had developed a thorough understanding of each item and were thus well-positioned to evaluate relevance. They were invited to evaluate the relevance of each item using a four-point Likert scale (1 = not relevant, 4 = highly relevant). The item-level content validity index (I-CVI, acceptable threshold > 0.78) and the average scale-level content validity index (S-CVI/Ave, acceptable threshold ≥ 0.90) were then calculated. Experts’ ratings resulted in I-CVI values ranging from 0.86 to 1.00 and an S-CVI/Ave of 0.996, indicating excellent content validity of the questionnaire.

Sample and data collection

This study was conducted at a tertiary-level pediatric hospital in Chongqing, China. The hospital is a National Clinical Research Center and Regional Children’s Medical Center with 2,480 beds, providing comprehensive pediatric services. Convenience sampling was used to recruit physicians and nurses from seven internal medicine and seven surgical departments. Inclusion criteria were: (a) holding a valid practicing certificate, (b) having at least one year of pediatric medical or nursing experience, and (c) providing informed consent. Exclusion criteria were: (i) non-hospital employees (e.g., trainees, visiting professionals), and (ii) staff on leave (e.g., sick leave, maternity leave).

According to the COSMIN checklist [15], the recommended sample size is at least five times the number of items. After two rounds of expert consultations and a pilot study, the final questionnaire comprised 36 items. With an additional 10% allowance for attrition, the minimum required sample size was 198 physicians and nurses.

The researchers coordinated with the head nurses of the participating departments and explained the procedures for questionnaire distribution and collection. The head nurses distributed the questionnaire link via departmental WeChat groups. Eligible physicians and nurses completed the electronic questionnaire, which consisted of a demographic form and the KAP questionnaire on pediatric pain management. Data collection was conducted from June 2 to June 14, 2024.

Participation was contingent upon informed consent, without which the survey closed automatically. The questionnaire was anonymous, and participants completed it independently. To ensure accurate understanding of the survey content, an introductory page was provided before the questionnaire, outlining the survey purpose, content, estimated completion time, and instructions for completion. To prevent duplicate submissions, each WeChat user was restricted to one response. In addition, all items were mandatory to ensure data completeness. Upon collection, two researchers reviewed the responses and excluded questionnaires with identical answers for all items, completion times shorter than 60 s, or inaccurate personal information.

Data analysis

Following data collection, systematic data screening was conducted to ensure dataset integrity. As no missing values were detected, missing data diagnostics and multiple imputation procedures were unnecessary. All statistical analyses were executed in R version 4.4.0 (R Foundation for Statistical Computing, Vienna, Austria) utilizing the tidyverse, psych, lavaan, semTools, and irr packages.

Sample characteristics and item distributions were summarized using descriptive statistics, including means (M), standard deviations (SD), and frequencies. Pearson’s chi-square test (with continuity correction/Fisher’s exact test for expected cell frequency < 5) was used for dichotomous/unordered categorical data, and the Mann-Whitney U test for ordinal ranked data, with a two-sided p < 0.05 considered statistically significant.

Prior to factor analysis, item analysis was performed to evaluate psychometric properties and guide initial item retention. Item discrimination was assessed using the Critical Ratio (CR) and the Corrected Item-Total Correlation (CITC). The CR was determined via independent samples t-tests comparing the upper and lower 27% of total score distributions; items with p < 0.05 and CR > 3.00 were retained. Furthermore, items exhibiting a CITC < 0.4 were deemed to have insufficient shared variance with the overarching construct and were marked for deletion.

To prevent overfitting and establish the generalizability of the factor structure, the total sample was randomly partitioned into two independent and approximately equal-sized subsamples. The first subsample was subjected to Exploratory Factor Analysis (EFA) to uncover the latent structure of the 36 items. The suitability of the data for EFA was confirmed via Bartlett’s test of sphericity (p < 0.05) and the Kaiser-Meyer-Olkin (KMO) measure, with values ≥ 0.80 indicating commendable sampling adequacy. Factor retention and item allocation were based on primary factor loadings ≥ 0.40. In instances of cross-loadings ≥ 0.40, items were assigned to the factor with the highest loading, supplemented by theoretical judgment.

Confirmatory Factor Analysis (CFA) employing maximum likelihood (ML) estimation was conducted on the second subsample to cross-validate the EFA-derived structure. Model fit was evaluated using the chi-square (χ²) statistic, Comparative Fit Index (CFI), Tucker-Lewis Index (TLI), Root Mean Square Error of Approximation (RMSEA), and Standardized Root Mean Square Residual (SRMR). CFI and TLI values > 0.90 were considered acceptable, with values > 0.95 indicating excellent fit. For RMSEA and SRMR, values < 0.08 were deemed acceptable, while those < 0.05 suggested good fit. Additionally, initial convergent validity for model refinement was estimated through the Average Variance Extracted (AVE), with values ≥ 0.50 considered adequate. Modification indices (MI > 10) were consulted to iteratively refine the model; for items exhibiting highly correlated residuals, the item with the weaker factor loading (λ) was removed. Only items with a standardized loading λ > 0.70 were retained in the final instrument.

Finally, to comprehensively establish the psychometric properties of the finalized instrument, reliability and ultimate validity analyses were executed using the total dataset. Scale reliability was established via internal consistency metrics, specifically Cronbach’s alpha (α) and McDonald’s omega (ω), with values > 0.60 considered acceptable. Validity was rigorously examined at multiple levels: overall construct validity was supported by the final CFA model; convergent validity was re-confirmed across the entire sample with AVE values ≥ 0.5; and discriminant validity was verified using the Fornell-Larcker criterion, ensuring that the square root of the AVE for each latent factor strictly exceeded its inter-factor correlations with other constructs.

Results

Sample

A total of 603 valid responses were included in the final analysis. The total sample was randomly partitioned into two independent, equally sized subsamples for exploratory factor analysis (EFA; n = 301) and confirmatory factor analysis (CFA; n = 302). Demographic variables—including age, gender, occupation, department, professional title, educational level, pain-related training experience, and years of working experience—were well-balanced between the two subsamples, with no statistically significant differences observed (all p > 0.05). Detailed demographic profiles are summarized in Table 3.

Table 3.

Demographic characteristics of the participants(N = 603)

Variables	EFA subsample(n = 301) n(%)	CFA subsample(n = 302) n(%)	χ2/Z	P
Age (years)			-1.81	0.07
< 30	67(22.26)	88(29.14)
30–39	189(62.79)	175(57.95)
≥ 40	45(14.95)	39(12.91)
Gender			0.06	0.81
Male	46(15.28)	44(14.57)
Female	255(84.72)	258(85.43)
Occupation			0.01	0.93
Nurse	252(83.72)	252(83.44)
Doctor	49(16.28)	50(16.56)
Department			0.02	0.89
Pediatric Internal Medicine	188(62.46)	187(61.92)
Pediatric Surgery	113(37.54)	115(38.08)
Title			-1.53	0.13
Deputy senior or above	30(10.00)	26(8.61)
Intermediate	132(43.85)	139(46.03)
Junior	139(46.18)	137(45.36)
Education level			-0.62	0.53
Doctorate	24(7.97)	16(5.30)
Master’s degree	33(10.96)	37(12.79)
Bachelor’s degree	240(79.73)	244(80.79)
Associate degree	4(1.33)	5(1.66)
Pain-related training			0.73	0.40
Yes	212(70.43)	203(67.22)
No	89(29.57)	99(32.78)
Working experience (years)			-0.93	0.35
< 7	90(29.90)	101(33.44)
7–14	150(49.83)	145(48.01)
≥ 15	61(20.27)	56(18.54)

Note. EFA = Exploratory Factor Analysis; CFA = Confirmatory Factor Analysis

Item analysis

Item discrimination was established using the critical ratio (CR) method. Independent samples t-tests revealed highly significant differences (p < 0.01) between the upper and lower 27% scoring groups across all items, with CR values ranging from 12.29 to 19.78. Corrected item-total correlations (CITC) ranged from 0.49 to 0.72, confirming strong internal homogeneity among the items. Based on these favorable psychometric metrics, all items were retained for subsequent factor analysis.

Exploratory factor analysis

EFA was performed on Subsample 1 (n = 301) using principal axis factoring with an oblique (Promax) rotation, reflecting the theoretical assumption that the underlying factors are correlated. The data demonstrated excellent suitability for factor extraction: the Kaiser-Meyer-Olkin (KMO) measure was 0.95 (well above the 0.80 threshold), and Bartlett’s test of sphericity was highly significant (χ2(630) = 12459.51, p < 0.001). Both parallel analysis and the scree plot converged on a 3-factor solution, which cumulatively accounted for 70.56% of the total variance. Factor loadings were robust, ranging from 0.57 to 0.93 (all ≥ 0.40), with no salient cross-loadings observed. The rotated factor structure is detailed in Table 4.

Table 4.

Exploratory factor analysis of the KAP questionnaire on pediatric pain management for physicians and nurses (n = 301)

No.	Description	Factor 1: Knowledge	Factor 2: Attitudes	Factor 3: Practices	h²
1.	Know the definition, classification, and mechanisms of pain.	0.69			0.56
2.	Know the characteristics and adverse effects of pediatric pain.	0.75			0.65
3.	Know the components of pediatric pain assessment.	0.78			0.70
4.	Know the timing and frequency of pediatric pain assessments.	0.77			0.71
5.	Know the commonly used pediatric pain assessment tools, their usage methods, target population, advantages, and disadvantages.	0.82			0.71
6.	Know the principles of preventive analgesia, multimodal analgesia, and stepwise analgesia for children.	0.80			0.69
7.	Know the commonly used non-pharmacological pain relief measures for children, including their applicable scope, advantages, and disadvantages.	0.85			0.79
8.	Know the implementation methods of commonly used non-pharmacological pain relief measures for children.	0.83			0.75
9.	Know the commonly used analgesic medications for children, including their indications and contraindications.	0.75			0.61
10.	Know the administration routes for commonly used pediatric analgesic medications.	0.74			0.61
11.	Know the common adverse reactions, prevention and management measures of pediatric analgesic medications.	0.76			0.65
12.	Know the methods for involving children and/or their families in pain management.	0.79			0.68
13.	Know the key points and requirements for medical/nursing records related to pediatric pain.	0.80		0.30	0.77
14.	Value effective pain relief for reducing children’s stress and promoting recovery.		0.88		0.85
15.	Consider the time, effort, and resources invested in pediatric pain management to be worthwhile.		0.91		0.88
16.	Recognize that timely, accurate, comprehensive, and dynamic pain assessment is critical for guiding rational clinical decisions.		0.90		0.87
17.	Value children’s pain complaints and/or related signs (e.g., heart rate, respiration, oxygen saturation, sweating, facial expressions, complexion, body movements, etc.).		0.89		0.87
18.	Support the principles of preventive analgesia, multimodal analgesia, and stepwise analgesia as effective approaches for reducing children’s pain and associated adverse reactions.		0.88		0.83
19.	Acknowledge the importance of assessing children’s and/or their families’ understanding of pain and providing relevant education.		0.90		0.88
20.	Consider involving children and/or their families in pain management to be essential.		0.90		0.87
21.	Believe that a multidisciplinary pain management team, including pediatrics, nursing, anesthesiology, pain medicine, pharmacy, and psychology, is essential for effective pain management.		0.91		0.90
22.	Value training in pediatric pain management for healthcare professionals as crucial for improving the quality of pain management.		0.93		0.91
23.	Support objective, truthful, accurate, timely, and comprehensive recording of the entire pain management process.		0.87		0.82
24.	Perform a comprehensive assessment of children’s pain, including its location, intensity, nature, duration, triggers, and associated complications.			0.66	0.56
25.	Choose appropriate pain assessment tools based on the children’s age, cognitive development level, and health status.	0.37		0.60	0.54
26.	Conduct regular pain assessments for children, such as upon admission, during painful procedures, postoperatively, and after pain interventions.	0.37	0.31	0.57	0.56
27.	Adjust pain assessment frequency based on the children’s needs and clinical condition, such as changes in pain intensity or the onset of new pain due to illness or activity.	0.37	0.30	0.66	0.65
28.	Minimize unnecessary painful procedures.		0.45	0.60	0.56
29.	Provide preventive analgesia for planned painful procedures, surgeries, or other foreseeable painful situations.			0.69	0.53
30.	Without affecting medical diagnosis, provide timely and proactive non-pharmacological and/or pharmacological pain relief for children based on pain assessment results.			0.74	0.65
31.	Manage adverse reactions related to pediatric pain interventions promptly and effectively.			0.78	0.71
32.	Consult a multidisciplinary pain management team for children with high pain scores and insufficient response to various pain interventions.			0.67	0.46
33.	Assess children’s and/or their families’ understanding of pain and provide relevant education.			0.77	0.66
34.	Encourage children and/or their families to participate in pain assessment and management.			0.82	0.73
35.	Record the entire pain management process objectively, truthfully, accurately, timely, and comprehensively.	0.30		0.75	0.69
36.	Follow up on children’s pain status after discharge and provide necessary advice or guidance.			0.73	0.55
	Eigenvalue	9.11	9.10	7.19
	% Variance explained	25.30%	25.28%	19.98%
	Cumulative % variance	25.30%	50.58%	70.56%

Note. h2 = Communalities. Extraction method: Principal Axis Factoring; Rotation method: Promax. The pattern matrix is reported. Primary factor loadings are displayed in bold. For clarity, factor loadings < 0.30 are suppressed

Confirmatory factor analysis

The EFA-derived structure was cross-validated using Subsample 2 (n = 302) via CFA with maximum likelihood (ML) estimation. Following initial model evaluation, iterative modifications led to the removal of four items (Items 28, 32, 33, and 36). Although theoretically conceptualized under the Practice dimension, these items exhibited substantial cross-loadings, evidenced by excessive modification indices (MIs) on the Attitude dimension (MI = 29, 24, 15, and 19, respectively), all clearly surpassing the predefined threshold of 10. Their exclusion markedly enhanced the standardized factor loadings within their target dimension and yielded a more parsimonious model. The finalized 3-factor model demonstrated an acceptable fit to the data: χ²(458) = 1873.66, p < 0.001, χ²/df = 4.10, CFI = 0.90, TLI = 0.90, RMSEA = 0.10 [95% CI: 0.08,0.11], RMR = 0.03, and SRMR = 0.06. All standardized factor loadings were statistically significant (p < 0.001), ranging from 0.68 to 0.95 (Table 5), completely fulfilling the recommended minimum criterion of 0.50.

Table 5.

Standardized factor loadings from confirmatory factor analysis of the KAP questionnaire on pediatric pain management (n = 302)

Factor	Item No.	Description	β	SE	p
Factor 1: Knowledge	1	Know the definition, classification, and mechanisms of pain.	0.69	-	-
	2	Know the characteristics and adverse effects of pediatric pain.	0.74	0.09	< 0.001
	3	Know the components of pediatric pain assessment.	0.72	0.10	< 0.001
	4	Know the timing and frequency of pediatric pain assessments.	0.75	0.10	< 0.001
	5	Know the commonly used pediatric pain assessment tools, their usage methods, target population, advantages, and disadvantages.	0.76	0.10	< 0.001
	6	Know the principles of preventive analgesia, multimodal analgesia, and stepwise analgesia for children.	0.79	0.10	< 0.001
	7	Know the commonly used non-pharmacological pain relief measures for children, including their applicable scope, advantages, and disadvantages.	0.85	0.10	< 0.001
	8	Know the implementation methods of commonly used non-pharmacological pain relief measures for children.	0.85	0.10	< 0.001
	9	Know the commonly used analgesic medications for children, including their indications and contraindications.	0.73	0.09	< 0.001
	10	Know the administration routes for commonly used pediatric analgesic medications.	0.74	0.09	< 0.001
	11	Know the common adverse reactions, prevention and management measures of pediatric analgesic medications.	0.83	0.09	< 0.001
	12	Know the methods for involving children and/or their families in pain management.	0.82	0.11	< 0.001
	13	Know the key points and requirements for medical/nursing records related to pediatric pain.	0.84	0.10	< 0.001
Factor 2: Attitudes	14	Value effective pain relief for reducing children’s stress and promoting recovery.	0.85	-	-
	15	Consider the time, effort, and resources invested in pediatric pain management to be worthwhile.	0.86	0.05	< 0.001
	16	Recognize that timely, accurate, comprehensive, and dynamic pain assessment is critical for guiding rational clinical decisions.	0.90	0.05	< 0.001
	17	Value children’s pain complaints and/or related signs (e.g., heart rate, respiration, oxygen saturation, sweating, facial expressions, complexion, body movements, etc.).	0.87	0.05	< 0.001
	18	Support the principles of preventive analgesia, multimodal analgesia, and stepwise analgesia as effective approaches for reducing children’s pain and associated adverse reactions.	0.91	0.05	< 0.001
	19	Acknowledge the importance of assessing children’s and/or their families’ understanding of pain and providing relevant education.	0.95	0.04	< 0.001
	20	Consider involving children and/or their families in pain management to be essential.	0.94	0.05	< 0.001
	21	Believe that a multidisciplinary pain management team, including pediatrics, nursing, anesthesiology, pain medicine, pharmacy, and psychology, is essential for effective pain management.	0.94	0.05	< 0.001
	22	Value training in pediatric pain management for healthcare professionals as crucial for improving the quality of pain management.	0.90	0.05	< 0.001
	23	Support objective, truthful, accurate, timely, and comprehensive recording of the entire pain management process.	0.91	0.05	< 0.001
Factor 3: Practice	24	Perform a comprehensive assessment of children’s pain, including its location, intensity, nature, duration, triggers, and associated complications.	0.75	-	-
	25	Choose appropriate pain assessment tools based on the children’s age, cognitive development level, and health status.	0.74	0.09	< 0.001
	26	Conduct regular pain assessments for children, such as upon admission, during painful procedures, postoperatively, and after pain interventions.	0.75	0.08	< 0.001
	27	Adjust pain assessment frequency based on the children’s needs and clinical condition, such as changes in pain intensity or the onset of new pain due to illness or activity.	0.76	0.08	< 0.001
	29	Provide preventive analgesia for planned painful procedures, surgeries, or other foreseeable painful situations.	0.73	0.09	< 0.001
	30	Without affecting medical diagnosis, provide timely and proactive non-pharmacological and/or pharmacological pain relief for children based on pain assessment results.	0.76	0.08	< 0.001
	31	Manage adverse reactions related to pediatric pain interventions promptly and effectively.	0.82	0.08	< 0.001
	34	Encourage children and/or their families to participate in pain assessment and management.	0.68	0.09	< 0.001
	35	Record the entire pain management process objectively, truthfully, accurately, timely, and comprehensively.	0.74	0.08	< 0.001

Note.β = Standardized factor loading; SE = Standard error. The first item in each factor serves as the reference indicator (SE and p not applicable). All factor loadings are statistically significant at p<0.001. Model estimation: Maximum likelihood. Software: lavaan package in R

Reliability & validity

Psychometric evaluation of the final instrument confirmed excellent reliability and validity. Internal consistency was highly satisfactory across all subscales, with Cronbach’s α ranging from 0.92 to 0.98 and McDonald’s ω ranging from 0.94 to 0.98 (Table 6). Convergent validity was robustly supported, as the Average Variance Extracted (AVE) values for all factors spanned from 0.56 to 0.84, safely exceeding the 0.50 threshold. Finally, discriminant validity was thoroughly established according to the Fornell-Larcker criterion, with the square root of the AVE for each latent construct strictly surpassing its corresponding inter-factor correlations. The final English and Chinese versions of the questionnaire are provided in the additional file 1.

Table 6.

Reliability and convergent validity indices of the finalized KAP questionnaire

Factor/Subscale	No. of Items	α	95% CI (α)	ω	95% CI (ω)	CR	AVE
Factor 1: Knowledge	13	0.96	0.95,0.96	0.96	0.95,0.96	0.96	0.63	0.79
Factor 2: Attitudes	10	0.98	0.97,0.99	0.98	0.97,0.99	0.98	0.84	0.92
Factor 3: Practice	9	0.92	0.90,0.94	0.94	0.92,0.96	0.92	0.56	0.75
Total Scale	32	0.96	0.95,0.97	0.97	0.96,0.98

Note. α = Cronbach’s alpha; ω = McDonald’s omega; CR = Composite Reliability; AVE = Average Variance Extracted; = square root of AVE. = Square root of AVE. The total scale composite reliability and AVE are conceptually specific to individual constructs and thus not computed for the multidimensional total scale

Discussion

Research indicates that insufficient pain knowledge among medical staff remains a significant barrier to effective pain management [17]. Additionally, outdated attitudes toward analgesia often impede the acquisition of current knowledge, hindering the practical implementation of effective pain management strategies [18]. The proactivity and standardization of medical staff’s practices directly determine the effectiveness of pediatric pain management. Understanding the knowledge, attitudes, and practices (KAP) of medical staff is essential for identifying gaps and informing improvements in pediatric pain care.

In this study, we developed and validated a KAP questionnaire to assess physicians’ and nurses’ knowledge, attitudes, and practices regarding pediatric pain management. The questionnaire was developed through a three-phase process, including item generation based on the KAP model and literature review, expert consultation, and a pilot survey to refine content. Psychometric testing showed excellent content validity (I-CVI: 0.86–1.00, S-CVI/Ave: 0.996). Exploratory factor analysis extracted three factors explaining 70.56% of the variance, and confirmatory factor analysis supported the three-factor structure. Convergent and discriminant validity were also established.

Compared with previous instruments that often focus on one or two aspects [10], this questionnaire integrates knowledge, attitudes, and practices into a single framework, enabling a more holistic assessment of pediatric pain management. Moreover, it also addresses key components of pain management such as prevention, assessment, intervention, evaluation, education, family involvement, and documentation, aligning with contemporary evidence in pediatric pain care [11, 19–21]. The present tool should be viewed as complementary to existing measures, offering broader coverage that could support both clinical audit and research on pain care delivery.

An additional feature of this questionnaire is its applicability across pediatric care settings. Whereas some tools have been developed for specific units (e.g., neonatal intensive care or perioperative contexts) [12, 14], our questionnaire covers core knowledge, attitudes, and practices relevant to pediatric pain management [11, 19–21]. This design supports its use among both physicians and nurses, across internal medicine and surgical contexts. Previous studies have highlighted that instruments with narrow applicability can hinder the generalization of results [22]. By addressing this limitation, the present questionnaire provides broader coverage that could facilitate consistent data collection and cross-setting comparisons, though further validation in different hospitals and cultural contexts is warranted.

The practical design of this questionnaire also supports its usability in real-world settings. Utilizing a 5-point Likert scale, the instrument can be completed in approximately six minutes, making it feasible for large-scale surveys, training evaluations, and routine clinical audits. Compared with test-based questionnaires [10, 13], self-report Likert instruments are generally more efficient and less burdensome for respondents, which promotes higher compliance and completion rates [23]. They may also reduce respondent stress and anxiety, encourage self-reflection, and improve response quality [24]. These features increase the practicality of the tool for integration into everyday clinical workflows. Nevertheless, as with other self-assessment tools, social desirability bias may exist, and responses may not fully reflect actual practices. Findings should be interpreted with this limitation in mind.

Limitations

As this was an initial development study, some limitations should be acknowledged and will need to be addressed in future research. Due to time and resource constraints, the study was conducted in a single hospital, and the demographic distribution of the sample was not fully balanced, which may have influenced the results. Although the questionnaire was designed for use by both physicians and nurses, the number of physicians included in both the Delphi panel and the psychometric evaluation was relatively small, which may have affected the results. Future studies should therefore include multicenter and more diverse samples with a larger proportion of physicians to further validate the questionnaire.

Conclusion

This study developed and preliminarily validated the Knowledge, Attitudes, and Practices (KAP) Questionnaire on Pediatric Pain Management for physicians and nurses. The final questionnaire comprises 32 items across three dimensions that cover key aspects of pediatric pain management. Psychometric testing demonstrated excellent content validity, high internal consistency, acceptable construct validity, as well as adequate convergent and discriminant validity. It provides a reliable and practical means to assess physicians’ and nurses’ knowledge, attitudes, and practices in pediatric pain management, and may help identify strengths and gaps in current practice and provide a basis for designing targeted training and quality improvement initiatives to enhance pediatric pain care.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

CR

Critical Ratio

CVI

Content Validity Index

CFA

Confirmatory Factor Analysis

EFA

Exploratory Factor Analysis

IAPS

The International Association for the Study of Pain

KAP

Knowledge, Attitudes, and Practices

PNKAS

The Pediatric Nurses’ Knowledge and Attitudes Survey

Author contributions

Q.S. conceptualized the study, developed the methodology, conducted the investigation, curated the data, performed formal analysis, wrote the original draft, and acquired funding. X.W. contributed to investigation and resources. W.L. contributed to methodology and resources. W.Z. contributed to investigation and data curation. H.L. contributed to conceptualization, validation, and supervision. X.Z. contributed to methodology, resources, supervision, funding acquisition, and project administration. All authors participated in reviewing and editing the manuscript, and read and approved the final version.

Funding

This work was supported by the Chinese Nursing Association Youth Research Project [ZHKYQ202410], the Chongqing Municipal Health Commission [2025WSJK002], and the National Natural Science Foundation of China [72074038].

Data availability

The datasets generated and/or analyzed during the current study are not publicly available due to ethical considerations but may be obtained from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

This study was conducted in accordance with the Declaration of Helsinki. Ethical approval was obtained from the Ethics Committee of the Children’s Hospital of Chongqing Medical University on May 27, 2024 (No. 209/2024). Participants gave informed consent to participate in the study before taking part.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.