Nursing and Midwifery Studies

: 2022  |  Volume : 11  |  Issue : 4  |  Page : 253--260

Development and psychometric assessment of barriers to postoperative pain management in dementia scale: A mixed methods exploratory study

Farzaneh Didvar1, Fatemeh Ghaffari2, Abbas Shamsalinia2,  
1 Student Research Committee, School of Nursing and Midwifery, Babol University of Medical Sciences, Babol, Iran
2 Nursing Care Research Center, School of Nursing and Midwifery, Babol University of Medical Sciences, Babol, Iran

Correspondence Address:
Abbas Shamsalinia
Nursing Care Research Center, School of Nursing and Midwifery, Babol University of Medical Sciences, Babol


Background: Identifying the barriers to acute pain management in older adults with dementia needs appropriate instruments adapted to the cultural structure of each community. Such an instrument can help to provide effective interventions. Objectives: This study aimed to develop and psychometrically validate Barriers to Postoperative Pain Management in Dementia Scale (BPPMDS) from the nurses' point of view. Methods: This methodological study was conducted in 2019–2020. In the first phase, semi-structured interviews with 15 nurses and literature review were conducted for item generation. In the second phase, face, content, and construct validity and reliability of the instrument were assessed. Results: In the first phase, 67 items were produced. In the second phase, the number of items reduced to 39 after the assessment of face, content, and construct validity. Exploratory factor analysis showed that the BPPMDS has three factors (i.e., older adult-related factors, health care provider-related factors, and system-related factors), which explain 57.57% of the total variance. The Cronbach's alpha coefficient of the scale was 0.956. Conclusions: The BPPMDS can be used as a valid and reliable scale to measure postoperative acute pain management in older patients with dementia and hip fracture.

How to cite this article:
Didvar F, Ghaffari F, Shamsalinia A. Development and psychometric assessment of barriers to postoperative pain management in dementia scale: A mixed methods exploratory study.Nurs Midwifery Stud 2022;11:253-260

How to cite this URL:
Didvar F, Ghaffari F, Shamsalinia A. Development and psychometric assessment of barriers to postoperative pain management in dementia scale: A mixed methods exploratory study. Nurs Midwifery Stud [serial online] 2022 [cited 2023 Mar 23 ];11:253-260
Available from:

Full Text


Hip fracture (HF) is a serious health problem in older adults. It is associated with pain and limitations in physical activity[1] and requires pain relief and rapid rehabilitation. Surgery is the most common treatment method in HF, and more than 98% of patients with HF undergo surgery.[2] Although surgery yields positive outcomes, it is accompanied by acute and severe pain, which affects the results of the intervention.[3] Older adults with dementia are at higher risk for inappropriate pain management after HF. The patient's cognitive impairment makes it difficult to verbalize pain. Furthermore, conflicting views of health-care providers about the side effects of medications in these vulnerable patients and organizational barriers such as staffing and equipment shortages make it difficult to assess and treat pain in these patients.[4] A valid instrument that fit the specific culture and structure of the healthcare system is needed to gather valid evidence and recognize the barriers to managing acute pain in older adults with dementia. Using such an instrument will improve the effectiveness of pain management interventions in older adults with dementia.[5],[6]

Formerly, an instrument was developed to assess the nurses' perceived barriers to optimal pain management in older adults. However, this instrument focused mainly on postoperative pain management and did not pass the standard psychometric evaluations.[7] Rantala et al. have also conducted a literature review to develop a specific instrument to assess the barriers to postoperative pain management in HF patients with dementia.[4] This questionnaire also has not been subjected to the standard psychometric assessments, the lived experiences[8] of patients and nurses have not been used in its development process, and it has not been translated and psychometrically tested in Iran or other countries.

Nurses working in postoperative acute care units often face barriers that limit their ability to manage acute pain in older adults with dementia. Since these nurses play the most important role in caring for the aforementioned patients, studying their opinions and experiences along with reviewing the literature would be an effective way to develop a valid and reliable instrument that fits the cultural conditions and structure of the Iranian healthcare system.[9],[8] Since no study has been conducted in this field and no valid and reliable instrument was available to measure the barriers to postoperative pain management in dementia patients with HFs, this study was conducted to fill the gap.


This study aimed to design and evaluate the psychometric properties of barriers to postoperative pain management in dementia scale (BPPMDS).


A sequential explanatory mixed methods study was conducted from June 2019 to January 2020. The study was conducted in two phases in the western regions of Mazandaran province (Ramsar, Tonekabon, Chalus, and Noshahr), Iran.

First phase: Item generation

Both inductive and deductive methods were applied for item generation. An explorative qualitative study was conducted to examine the nurses' experiences and inductively generate an item pool. Nurses who had experience of caring for older adults with dementia and HFs were purposively invited to participate in the study. Finally, 15 eligible nurses participated in this phase. Semi-structured interviews were conducted to gather the participants' experiences and perceptions of barriers to postoperative pain management in patients with dementia who underwent surgery for HF. Graneheim and Lundman's approach was used to analyze the qualitative data. Accordingly, all interviews were transcribed word by word and the transcripts were read through several times to get an overall impression of the whole texts. Meaning units were identified and condensed into codes. Similar codes were grouped and then clustered into subcategories. Similar subcategories were grouped together and then clustered into categories through the constant comparison method. This process was repeatedly checked and reviewed by the research team members. The categories were labelled afterwards. The codes emerged in this phase were used to generate the initial item pool.[10] The trustworthiness of the data was checked using Guba and Lincoln's four criteria of credibility, dependability, conformability, and verifiability.[11]

At the next step, we systematically reviewed the SCOPUS, PUBMED, PsycINFO, and Web of Science databases to find studies on barriers to postoperative pain management in dementia to deductively develop the generated item pool. Totally, eight relevant English-language documents published between 2000 and 2020 were found, coded, and the resulting codes added to the primary item pool.

Integration of qualitative and literature reviews phases

The codes and items extracted from the literature were added to those that emerged from the qualitative study, and a common item pool was created. The item pool was rechecked by the research team, duplicates were removed, similars were merged, and some items were modified. Finally, 67 items remained in the first draft of the instrument.

Second phase: Psychometric evaluation

The initial draft of the instrument was then modified and developed through face, content, construct, convergent and divergent validity, and reliability assessment.

Face validity assessment

Face validity was measured considering both the qualitative and quantitative approaches. To this end, 10 nurses were asked to share their opinions on the readability, wording, difficulty, relevance, and ambiguity of each item. Their suggestions were applied to the related items. The same nurses were then asked to comment on the importance of each item using a five-point Likert scale ranging from “5: absolutely important” to “1: not important.” After calculating the mean importance of each item and the frequency of nurses who rated an item as 4 or 5, we calculated the impact scores of the items using the following formula: Item Impact Score = Importance (%) × Frequency

Content validity assessment

In the qualitative content validity assessment, 10 experts in nursing education, psychometry, qualitative studies, and clinical job assignment were asked to review the scale and share their recommendations. Their comments were applied. Then, the experts were invited to rate each item's essentiality (as essential, useful but not essential, or unessential) and relevance (as not relevant, somewhat relevant, quite relevant, and very relevant). Then, the content validity index (CVI) of each item was determined based on Lawshe's table (1975).[12] Accordingly, items with a CVI ≥0.62 were accepted. The CVI was calculated by dividing the number of experts who rated an item as 3 or 4 by their total number.[13] Items with a CVI >0.79 were accepted.[14]

Construct validity assessment

Exploratory factor analysis (EFA) and confirmatory factor analysis (CFA) were conducted to assess the construct validity of the BPPMDS. A cross-sectional study was carried out on a convenient sample of 450 nurses, of whom 230 and 220 ones were used for the EFA and CFA, respectively. Inclusion criteria included at least a bachelor's degree in nursing, work experience in surgical wards and intensive care units, and experience of caring for older adults with dementia after HF surgery [Table 1].{Table 1}

The Kaiser-Meyer-Olkin (KMO) test for sample adequacy was used to determine the fitness of the data. Bartlett's test of sphericity was then applied to ensure that the items in the instrument were sufficiently correlated. Factors with eigenvalues ≥1 were considered significant. Latent factors were extracted using Principal Axis Factoring, Varimax Rotation, and Scree Plot. Item communalities <0.4 were removed. Then, the extracted factors were evaluated through CFA. Acceptable model fit is indicated by: Parsimonious comparative fit index (PCFI) and Parsimonious normed fit index (PNFI) (>0.5), comparative fit index (CFI) and incremental fit index (IFI) (>0.9), Root Mean Square Error of Approximation (RMSEA) (<0.08), and Chi-square/degree-of-freedom ratio (χ2/df) (<3 good, <5 acceptable).[15]

Convergent and divergent validity assessment

The convergent and divergent validity of the BPPMDS were evaluated based on the Fornell and Larcker approach. The following criteria must be satisfied to ensure convergent validity: Composite reliability (CR) >0.7, average variance extracted (AVE) >0.5, and CR > AVE. Moreover, the maximum shared squared variance (MSV) and Average Shared Square Variance (ASV) must be less than AVE to confirm divergent validity.[16] The significance level was set at P < 0.05 for all tests.

The normal distribution of the data was tested by handling the single-and multivariable distribution of the data and the outlier data separately. The multivariable outlier data were analyzed using the Mahalanobis d-squared (P < 0.001) and Mardia's coefficient of multivariate kurtosis >8.[17] The percentage of missing data was assessed using the multiple imputation method and was then replaced by the mean response given by the participants.

Reliability assessment

The reliability of the BPPMDS was evaluated using its internal consistency and stability. Internal consistency was assessed by calculating Cronbach's alpha and McDonald's omega coefficients in a sample of 450 nurses. A reliability >0.7 was considered acceptable. For stability assessment, the scale was completed by 30 nurses with a 2-week interval.

Standard error of measurement

The value of standard error of measurement (SEM) for all subscales and the whole scale was measured, and the Minimal Detectable Change (MDC) and the minimal important change (MIC) were compared to determine final agreement.


After determining the weight of each item, the standard 0–100 scoring scale was used. The linear transformation formula was applied for convert the scores into a 0–100 scale[18] as shown in the following formula:

x = actual raw score − lowest possible raw score * 100/possible raw score range

Statistical analysis

The EFA and statistical tests were conducted in SPSS software v. 24.0 (Armonk, NY: IBM Corp), CFA in AMOS v. 24.0, and McDonald's omega coefficient in JASP.

Ethical considerations

This study received approval from the Ethics Committee of Babol University of Medical Sciences (IR.MUBABOL.HRI.REC.1398.097). Participants were explained about the study protocols and aims and their written consent was obtained. They were assured about the confidentiality of their information. In the qualitative phase of the study, with the permission of the participants, their voices were recorded. Names of the participants were not mentioned in the study; instead, codes were used in the interview texts.


The results show that most of the participants were female (87/9%). The mean age of the samples was 36.42 ± 7.76 years. Most of the samples were married (74.2%) and had a bachelor's degree (89.4%). The highest work experience was 5–10 (26.1%) and worked in the surgery department (34.5%). The most of samples had no history of acute pain (87%) and no history of surgery (66.4%) [Table 1].

The outcome of the item generation phase

At the end of the item generation phase, an item pool of 75 items was created, 29 of which came from the qualitative phase and 46 from the literature. After merging some items, 67 items entered the psychometric assessments.

Outcome of the psychometric evaluation

Face and content validity

Thirteen items were removed due to having an impact score of ≤1.5. Three items were revised after the qualitative content validity assessment. Nine items were removed for having a CVR <0.62 and 4 for having a CVI <0.79. Finally, 41items remained for construct validity assessment.

Construct validity

The sampling adequacy index was 0.879, and the results of Bartlett's test were significant (df = 741; x2 = 9148.396; P ≤ 0.001). After applying the Principal Axis Factoring and varimax rotation, three factors with eigenvalues >1 were extracted. These factors were labeled “older adult-related factors,” “health care provider-related factors” and “system-related factors.”

The three extracted factors cover 57.572% of the total variance of pain management barriers among older adults with dementia after HF surgery [Table 2]. After varimax rotation, the items of “nurses' reluctance to administer prescribed painkillers to older patients (especially those with dementia or delirium) due to fear of overdose” and “attitude that older adults would die anyway” were removed due to factor loading <0.4.{Table 2}

In CFA, the initial model was not fit, however, the Chi-square goodness-of-fit was adequate after model justification and drawing the correlation between the measured errors. We then calculated other indices (PCFI = 0.836, PNFI = 0.78, CMIN/DF = 1.959, RMSEA = 0.064, IFI = 0.923, and CFI = 0.901), and all confirmed the fitness of the final model. According to the final model of the factorial structure of BPPMDS, correlation was found between the measurement errors of e4/e6, e14/e16, e19/e25, and e33/e35 [Table 3] and [Figure 1].{Figure 1}{Table 3}

Convergent and divergent validity

The AVE of all factors was >0.5 (0.51–0.70). The AVE of each factor was also greater than its ASV (0.41–0.48) and MSV (0.4–0.51). The BPPMDS construct showed acceptable convergent and divergent validity [Table 4].{Table 4}


The Cronbach's alpha and McDonald's omega coefficients were >0.7 for all BPPMDS items, and the overall Cronbach's alpha coefficient was 0.656. The scores achieved in the test retest were compared using the Intraclass Correlation Coefficient (ICC), and the results showed a significant correlation (P < 0.001). The overall ICC index of the whole instrument was 0.923 [Table 5].{Table 5}

Regarding the SEM, the results showed that the value of MDC was greater than MIC [Table 6].{Table 6}


The final edition of the BPPMDS includes 39 items in three factors of elderly-related factors (10 items), healthcare provider-related factors (25 items), and system-related factors (4 items). A five-point Likert scale was used to quantify the items (5: highly agree, 4: agree, 3: no comments, 2: disagree, and 1: highly disagree). The total scale score ranges from 39 to 195.


In this study, inductive and deductive methods were used for item generation. The combination of these methods is introduced as the most appropriate technique for this purpose.[19] In addition to the fact that the nurses confirmed the wording and clarity of the items, the impact scores of all items were >1.5, indicating the face validity of the BPPMDS items.[20] Furthermore, the experts panel reviewed the grammar, wording, item allocation, and scaling of the instrument. The experts' comments were also used to calculate the CVR and CVI of the BPPMDS, and the results showed that the scale has acceptable CVR and CVI. It is believed that two points need to be considered when examining content validity: The first is to ensure that the most relevant and appropriate items are selected, and the second is to design the items in the most appropriate form. The former is referred to as CVR and the latter as CVI.[21]

We conducted EFA and CFA to examine the construct validity of the BPPMDS. Factor analysis is a well-known method for categorizing the items into factors (subscales).[22] The results of KMO and Bartlett's tests showed the adequacy of the sample for conduction factor analysis. KMO values of 0.7–0.8 are considered adequate and 0.8–0.9 as highly adequate.[23] The results of CFA (i.e., RMSEAR, CFI, NFI, and Chi-squared values) also confirmed the fitness of the model.[21],[24]

The results show good internal consistency or reliability based on the acceptable value of Cronbach's alpha for the whole instrument. Conventionally, this coefficient must be >0.7.[25] The stability of the scale was also measured through test-retest and ICC methods. A previous study confirmed these methods for measuring the reliability of a scale.[26] The strengths of this study include the relatively large sample size, sampling from different hospitals in different cities, the use of a sequential, exploratory mixed-methods design, and inductive and deductive methods to generate items. However, there are some limitations in this study. The BPPMDS is a self-report scale and has all the limitations of a self-report instrument. The geographically-limited sample is itself a limitation, as social, cultural, and regional factors may influence participants' experiences and therefore, the formation of the scale items. Hence, further studies are needed to confirm the validity and reliability of this instrument in different cultural contexts. Similar studies are also suggested to examine how older patients view barriers to postoperative acute pain management.


The BPPMDS is a valid and reliable 39-item instrument which includes three subscales: older adults-related factors, health care providers-related factors, and system-related factors. This scale can be used as a valid and reliable instrument to measure postoperative acute pain management in older adults after any fracture or surgery because the items designed are not specific to patients with HFs. Nursing managers can use the BPPMDS to understand barriers to pain management in acute care units.


We appreciate all the participants who graciously gave their time to answer the questions and helped us in this study.

Financial support and sponsorship

This article is derived from a thesis research project supported by the research deputy at Babol University of Medical Sciences.

Conflicts of interest

There are no conflicts of interest.


1Turunen K, Salpakoski A, Edgren J, Törmäkangas T, Arkela M, Kallinen M, et al. Physical activity after a hip fracture: Effect of a multicomponent home-based rehabilitation program – A secondary analysis of a randomized controlled trial. Arch Phys Med Rehabil 2017;98:981-8.
2Membership of the Working Party, Griffiths R, Alper J, Beckingsale A, Goldhill D, Heyburn G, et al. Management of proximal femoral fractures 2011: Association of Anaesthetists of Great Britain and Ireland. Anaesthesia 2012;67:85-98.
3Zanker J, Duque G. Rapid geriatric assessment of hip fracture. Clin Geriatr Med 2017;33:369-82.
4Rantala M, Kankkunen P, Kvist T, Hartikainen S. Barriers to postoperative pain management in hip fracture patients with dementia as evaluated by nursing staff. Pain Manag Nurs 2014;15:208-19.
5Achterberg W, Lautenbacher S, Husebo B, Erdal A, Herr K. Pain in dementia. Pain Rep 2020;5:e803.
6Waltz CF, Strickland OL, Lenz ER. Measurement in Nursing and Health Research. New York: Springer Publishing Company; 2010.
7Midtbust MH, Alnes RE, Gjengedal E, Lykkeslet E. Perceived barriers and facilitators in providing palliative care for people with severe dementia: The healthcare professionals' experiences. BMC Health Serv Res 2018;18:709.
8Collins KM, Onwuegbuzie AJ, Sutton IL. A model incorporating the rationale and purpose for conducting mixed methods research in special education and beyond. Learn Disabil 2006;4:67-100.
9Bahrami M, Bidgoli Golkhatmi M, Saadati M, Saadati A, Barati M, Zareayan Jahromi N, et al. A study of patients and nurses' perception of pain management after cardiac surgery. J Health Care 2016;18:179-90.
10Graneheim UH, Lundman B. Qualitative content analysis in nursing research: Concepts, procedures and measures to achieve trustworthiness. Nurse Educ Today 2004;24:105-12.
11Lincoln YS, Guba EG. But is it rigorous? Trustworthiness and authenticity in naturalistic evaluation. New Directions for Program Evaluation 1986;30:73-84.
12Lawshe CH. A quantitative approach to content validity. Person Psychol 1975;28:563-75.
13Sharma S. Nursing Research and Statistics. New Delhi: Elsevier Health Sciences; 2018.
14Almanasreh E, Moles R, Chen TF. Evaluation of methods used for estimating content validity. Res Social Adm Pharm 2019;15:214-21.
15Collier JE. Applied Structural Equation Modeling Using AMOS: Basic to Advanced Techniques. London: Routledge; 2020.
16Fornell C, Larcker DF. Evaluating structural equation models with unobservable variables and measurement error. J Market Res 1981;18:39-50.
17Black W, Babin BJ. Multivariate data analysis: Its approach, evolution, and impact. In: Babin B, Sarstedt M, editors. The Great Facilitator. Cham: Springer; 2019. p. 121-30.
18White MK, Maher SM, Rizio AA, Bjorner JB. A meta-analytic review of measurement equivalence study findings of the SF-36® and SF-12® Health Surveys across electronic modes compared to paper administration. Qual Life Res 2018;27:1757-67.
19Boateng GO, Neilands TB, Frongillo EA, Melgar-Quiñonez HR, Young SL. Best practices for developing and validating scales for health, social, and behavioral research: A primer. Front Public Health 2018;6:149.
20Timkova V, Nagyova I, Reijneveld SA, Tkacova R, van Dijk JP, Bültmann U. Quality of life of obstructive sleep apnoea patients receiving continuous positive airway pressure treatment: A systematic review and meta-analysis. Heart Lung 2020;49:10-24.
21Kile D, Eaton M, deValpine M, Gilbert R. The effectiveness of education and cognitive rehearsal in managing nurse-to-nurse incivility: A pilot study. J Nurs Manag 2019;27:543-52.
22Kazim A, Shankar A, Jindal M. Statistical methods and applications: A comprehensive reference for the healthcare industry. In: Exploratory Data Analytics for Healthcare. 1st ed. Florida: CRC Press; 2021.
23Sharif Nia H, Ebadi A, Lehto RH, Mousavi B, Peyrovi H, Chan YH. Reliability and validity of the Persian version of templer death anxiety scale-extended in veterans of Iran-Iraq warfare. Iran J Psychiatry Behav Sci 2014;8:29-37.
24Braviano G, Scolari S. Multivariate analysis in design. J Geom Graph 2022;26:147-57.
25Harrison CJ, Sidey-Gibbons CJ. Modern psychometric measurement and computerized adaptive testing. In: Handbook of Quality of Life in Cancer. Cham: Springer; 2022.
26Buchholz I, Janssen MF, Kohlmann T, Feng YS. A systematic review of studies comparing the measurement properties of the three-level and five-level versions of the EQ-5D. Pharmacoeconomics 2018;36:645-61.