It is said that quality on health care is a difficult concept to define and measure 1. Although, as an object of study it has been developed and studied in the last decades, interesting researchers to measure it and use it in health policy evaluation scenarios.2. The definition of quality applied to nursing care is multidimensional and complex 3. Research approaches the concept but without delving into it, limit its meaning to value judgments of the authors. Quality is then conceived as a degree of conformity with current standards 4, which ends in multiple ways of seeing it from each of the forms presented by the measurement instruments used to evaluate it. A professional and comprehensive nursing approach to its study defines quality as part of the characteristics of the environment in which nursing care is provided, and the provision of the service; These include the types of processes and activities that nursing staff perform in pro-viding care to patients 5.
The interest in quality measurement is supported by technological and scientific advances to achieve better measurements including robust data analysis software, standardized methods and techniques 6.
The perception of the satisfaction of individuals and their value judgment on quality has been taken to the hospital context, allowing the health care provided by nurses to be evaluated, with the aim of improving both the service provided and the professional nursing practice 7-9. The Latin root "sat is", which means "sufficient", gives us the etymology of "satisfaction", deriving the adequate fulfilment that satisfies expectations, needs or desires and, making it a priority to give what is required, leaves no room for complaint 6. A patient satisfied with the care offered improves the effectiveness of the care, allows better results with the medication, the advice to maintain their health and reduces the probability of readmission 10. Measuring patient satisfaction with nursing care is important to assess and meet patient needs and to determine appropriate nursing interventions. The nursing staffs, being the largest payroll in hospital care, give a higher percentage to the perception of the global quality of care.
Evaluating the satisfaction of patients with the care received and with health care allows correcting interventions that will not only directly improve medical care and the patient's condition, but, at the same time, increase patient satisfaction, which will lead to a positive response to treatment. This becomes important to reward and boost staff morale. Patient experience is being widely considered in the evaluation of healthcare service quality 11. Thus this study aims to test the validity, reliability of the CUCASE (Quality of care questionnaire of nursing), in Colombia.
METHODS
Participants for the facial and content validity
The CUCACE was adapted from Castilian (Spain) to Spanish (Colombia), translated and adapted locally. The content validity was determined by a panel of experts. Each expert was given a spreadsheet, they reviewed and rated the relevance of the 26 Experiences Nursing Care Scale (ENCS) items and 19 Satisfaction with Nursing Care Scale (SNCS) items, using a four-point rating scale ranging from 1 (not important) to 4 (very important). The group of experts was made up of 14 nursing professionals, specialized in health administration, community nursing, public health, and auditing. The Content Validity Index (CVI) of each component was calculated based on the rating by experts.
Participants for the construct validity and exploratory factor analysis and Reliability
A sample of 261 patients in two public hospitals of Boyacá (Colombia) was recruited on the day of discharge, from June 2018 to May 2019. The criteria for selecting candidates consisted of being patients with more than 48 hours of stay. Patients who denied participating in the study were excluded. According to literature, this sample size met the criteria for an adequate validity and reliability analyses, considering at least 10 subjects per items 12-15.
Instrument
The CUCACE is a Spanish version of the Newcastle Questionnaire Satisfaction with Nursing Scales (NSNS) validated by 16. The instrument has a perceptible language, regardless of the cultural level of each surveyed; its application takes approximately 15 minutes. The NSNS was developed by Thomas et al.17.
It measures the experiences and satisfaction of patients with the nursing care received, depending on their perspective. The questionnaire is self-completed by the patients and they judge the quality of care as good or bad. They evaluate the concepts of availability and the care of nurses, the individual treatment provided, the provision of security and information, the openness of the informality of nurses, the professionalism and knowledge capacity of the nurses, the organization and environment of the room 18.
The CUCACE is made up of three sections: (i) Nursing Care Experiences Scale, (ii) Satisfaction with Nursing Care and (iii) demographic information section 16,17,19.
The ENCS is made up of 26 statements that evaluate aspects of the nursing staffs experience, using a seven-point Likert scale (1=completely disagree; 7=completely agree).
There are 11 questions with negative statements that avoid statement bias. The items are recoded; add all the answers and we get an overall experience score, with a potential range of 0 to 100 16,20,21.
The SNCS is built with 19 items. Each item is scored on a five-point Likert scale (1=not satisfied at all; 5 completely satisfied). The sum of the transnphromated responses is called general satisfaction score, which is also from 0 to 100 16,20,21. In the study by Thomas et al., Cronbach's alpha was 0.89 for ENCS and 0.96 for SNCS. The correlations between the individual items and the total ranged from 0.53 to 0.82 18,19,21.
The final section asks about patient demographic information and details of the hospital stay. This section also includes a one-item scale (seven-point response scale) on the overall satisfaction of patients with their recent hospital stay.
Data collection and ethical considerations
To collect the data, the manual for the use of CUCASE pro-vided by the Spanish research group is followed verbatim, who also gave the permits and advice for its use 16. The study was approved by the hospital administration and ethical extern committee. The Researches Ethics Committee of UPTC Nursing Graduate Association (www.agenf.org) gave approval for the study (Act 01: 2018/01/10).
After explaining the objectives of the study, the safe and confidential handling of personal data; each of the patients was invited to participate in the study, they voluntarily accepted and signed a consent form.
Data analysis
Following Lawshe's model for the quantitative assessment of content validity modified by Tristan 22, each item was evaluated through the measurement of the Content Validity Ratio (RVC), that determines whether an item is essential to evaluate the construct. It is useful, but dispensable or unnecessary. This formula assigns a score between -1 and +1, where a value equal to or greater than 0.51 is sufficient to leave the item in the final version of the instrument. Finally, once those elements that have values higher than the minimums proposed by Lawshe have been defined, the average CVR is calculated to obtain the CVI of the entire test, whose value must be greater than 0.8.
The construct validity was made by exploratory factor analysis, previously an evaluation was made by means of sample adequacy (KMO), proposed by Kaiser, Meyer and Olkin; and Bartlett's test of sphericity. The Varimax method was used for the rotation of the components, which minimizes the number of variables with high loads by one factor, improving the interpretation capacity, accepting those with Pearson's R> 0.4. Factorial analysis that is a multivariate statistical technique that serves to study the dimensions that underlie the relationships between variables 13,15,23.
Reliability was carried out by evaluating the homogeneity of the different items of the scale by calculating the internal consistency for each item individually. For this purpose, Alfa of Cronbach and McDonald's W 13-15,23,24.
RESULTS
Table 1 shows the result of facial validity and the calculation of the Content Validity Index of the ENCS and the SNCS. The items on the scales obtained values greater than 0.80; which indicates that they should be included in the final version of the validated instrument. Similarly, a global CVI of the scale is found with a value of 0.88 for the ENCS and 0.94 for the SNCS, which confirms the essentiality of the instrument.
The average of the score given to the properties mentioned above in the items was higher than 0.78 in the two scales (Table 1), considering the items as valid for facial assessment. The nurse professional expert considers that the subscales ENSC and SNSC were satisfactory in clear, precision and understanding.
Construct validity by subscales ENCS and SCNS
The mean age of participants was 54 ± 16 (with a range from 18 to 81) years. Half of the participants (52.5%) were women and most of them were married (81%). The statistics descriptive by two scales ENCS and SNCS (Table 2)
In the patient's sample (n=261). By subscale ENCS, KMO was 0.72. Barlett's test of sphericity was significant (X2=7211.040. d.f=325. p<0.001).
N | Min | Max | Mean | Sd | Variance | |
---|---|---|---|---|---|---|
Total ENCS | 261 | 55,13 | 97,44 | 88,07 | 9,86 | 97,19 |
Total SNCS | 261 | 34,21 | 84,21 | 64,18 | 10,00 | 100 |
Seven factors had initial eigenvalue over one. Explaining 79.77% of the variance. The PCA with Varimax rotation resulted too in the seven factors extraction solution. Which is presented in the Table 2. The first factor explaining 17.77% of the total variance, include the items [4. 20. 5. 3. 24. 6] The second factor explaining 14.05% of the total variance, include the items [18. 11. 7. 1] The three factor explaining 11.15% of the total variance, include the items [14. 23. 9. 8] The fourth factor explaining 10.91% of the total variance, include the items [2. 16. 10] The five explaining 9.16% of the total variance include the items [21. 15. 13] The six factor explaining 8.38% of the total variance, include the items [17. 26. 12] The seven factor explaining 8.34% of the total variance; include the items [25. 19. 22] unique with negative value.
In the patients sample (n=261). By subscale SNCS, KMO was 0.808 and Barlett's test of sphericity was significant (X2=1749.176. d.f=171. p<0.01) (Table 3).
Items | Components | ||||||
---|---|---|---|---|---|---|---|
1 | 2 | 3 | 4 | 5 | 6 | 7 | |
4 | 0.96 | -0.079 | 0.111 | -0.09 | 0.036 | 0.027 | -0.001 |
20 | 0.925 | -0.048 | 0.263 | -0.072 | 0.039 | 0.01 | 0.013 |
5 | 0.816 | 0.264 | 0.068 | 0.205 | -0.047 | -0.072 | -0.061 |
3 | 0.641 | 0.451 | -0.068 | 0.509 | 0.062 | -0.124 | 0.006 |
24 | 0.587 | 0.328 | -0.173 | 0.342 | 0.384 | 0.276 | 0.187 |
6 | 0.528 | 0.435 | -0.086 | 0.422 | 0.161 | 0.335 | -0.056 |
18 | -0.002 | 0.902 | -0.145 | 0.248 | 0.07 | 0.042 | 0.026 |
11 | -0.043 | 0.739 | 0.048 | 0.039 | 0.102 | 0.329 | 0.214 |
7 | 0.49 | 0.704 | 0.176 | 0.143 | 0.011 | -0.122 | 0.089 |
1 | 0.307 | 0.542 | 0.322 | 0.152 | 0.258 | 0.343 | 0.038 |
14 | 0.193 | 0.072 | 0.833 | 0.09 | 0.12 | -0.029 | 0.068 |
23 | -0.029 | -0.095 | 0.711 | -0.075 | -0.102 | 0.271 | -0.114 |
9 | 0.049 | -0.004 | 0.699 | 0.035 | 0.101 | -0.081 | -0.005 |
8 | 0.61 | 0.124 | 0.62 | 0.075 | 0.086 | 0.013 | 0.168 |
2 | 0.063 | -0.005 | 0.1 | 0.927 | -0.008 | 0.231 | 0.065 |
16 | 0.032 | 0.377 | -0.008 | 0.862 | -0.064 | -0.08 | -0.005 |
10 | 0.108 | 0.437 | 0.384 | 0.493 | 0.112 | 0.367 | 0.215 |
21 | -0.078 | 0.082 | 0.13 | 0.043 | 0.872 | 0.017 | -0.038 |
15 | 0.01 | 0.365 | 0.162 | -0.072 | 0.828 | -0.016 | 0.007 |
13 | 0.444 | -0.237 | -0.054 | -0.021 | 0.693 | 0.141 | 0.013 |
17 | -0.069 | 0.084 | 0.011 | 0.086 | 0.011 | 0.912 | 0.069 |
26 | 0.084 | 0.289 | 0.498 | 0.043 | 0.166 | 0.509 | 0.232 |
12 | 0.129 | 0.447 | 0.101 | 0.389 | 0.04 | 0.487 | 0.267 |
25 | -0.041 | 0.082 | 0.13 | -0.091 | -0.025 | 0.05 | 0.912 |
19 | -0.077 | 0.246 | 0.015 | 0.124 | -0.13 | 0.283 | 0.784 |
22 | -0.374 | 0.089 | 0.246 | -0.251 | -0.294 | 0.142 | -0.631 |
Six factors had initial eigenvalue over one explaining 65.11% of the variance. The PCA with Varimax rotation resulted in the six factors extraction solution. Which is presented in the Table 4. The first factor explaining 20.60% of the total variance, include the items [7. 13. 11. 10. 9. 14] The second factor explaining 10.40% of the total variance, include the items [17. 12. 1. 8] The three factor explaining 9.44% of the total variance, include the items [16. 19. 4] The fourth factor explaining 9.29% of the total variance, include the items [3. 15. 5] The five explaining 8.26% of the total variance include the items [18. 6] The six factor explaining 7.11% of the total variance, include the item [2].
Item | Components | |||||
---|---|---|---|---|---|---|
1 | 2 | 3 | 4 | 5 | 6 | |
7 | 0.82 | 0.013 | 0.189 | 0.122 | 0.048 | -0.029 |
13 | 0.763 | 0.223 | 0.159 | 0.069 | 0.206 | 0.127 |
11 | 0.71 | 0.044 | 0.008 | 0.257 | 0.247 | 0.007 |
10 | 0.701 | 0.26 | 0.052 | 0.312 | -0.175 | 0.097 |
9 | 0.648 | 0.121 | 0.394 | 0.014 | 0.253 | 0.005 |
14 | 0.482 | -0.003 | 0.435 | -0.017 | 0.458 | 0.208 |
17 | 0.167 | 0.726 | -0.073 | 0.465 | -0.019 | -0.072 |
12 | 0.012 | 0.69 | 0.108 | 0.019 | 0.132 | 0.226 |
1 | 0.419 | 0.6 | 0.206 | -0.136 | 0.054 | -0.175 |
8 | 0.398 | 0.493 | -0.199 | -0.131 | -0.014 | 0.397 |
16 | 0.24 | 0.005 | 0.774 | 0.119 | -0.078 | -0.017 |
19 | 0.156 | 0.118 | 0.418 | 0.254 | 0.253 | -0.118 |
4 | -0.364 | 0.372 | 0.39 | 0.191 | 0.045 | 0.378 |
3 | 0.066 | 0.026 | 0.25 | 0.745 | 0.28 | -0.062 |
15 | 0.481 | 0.051 | -0.035 | 0.566 | 0.059 | 0.181 |
5 | 0.377 | 0.075 | 0.398 | 0.557 | -0.095 | 0.267 |
18 | 0.219 | -0.044 | -0.193 | 0.122 | 0.791 | 0.043 |
6 | 0.014 | 0.259 | 0.319 | 0.101 | 0.589 | 0.003 |
2 | 0.1 | 0.074 | -0.019 | 0.064 | 0.05 | 0.877 |
Reliability
A Cronbach's alpha was obtained that ranged from 0.844 to 0.867 for each item of the SNCS. The final scale alpha is 0.863 and McDonald's w 0.889.
The internal consistency of the ENCS. Using Cronbach's alpha, it ranged from 0.822 to 0.885 for the ENCS items. The Cronbach of the SNCS global satisfaction scale is 0.847 and the McDonald's w is 0.872.
DISCUSSION
Although the version of the CUCACE has only been partially validated in Spain and used in Colombia, due to its origin in the NSNS, we compare its psychometric proper-ties with the validated versions in other countries.
The results of the present study reaffirm the internal consistency measured by the Cronbach alpha in the CU-CACE of 0.851 and 0.863, for ENCS and 0.847 for SNCS, being very similar to those reported in Spain 16. and somewhat lower than the original English instrument 19.
The NSNS has been validated for different populations; it is common to find only the alphas internal consistency report between 0.75 to 0.95 for ENCS and from 0.93 to 0.98 for SNCS 25-28; and it has also been translated into different languages, using it in various countries such as Jordan 25, Italy 27,28, Canada 26, England 29, Poland 30,31, Ethiopia 32 or Brazil 33,34.
The NSNS has also been translated and validated into Turkish (Istanbul) in the context of medical and surgical patients 18 in a sample of 200 patients obtain an alpha of 0.96 for the total scale and for the items from 0.43 to 0.89. 35 in the same context, with a sample of 229 medical and surgical patients, reported an internal consistency of 0.95 for the full scale.
Torres used the CUACASE 36, the version translated and validated into Spain 16, in a sample of 180 patients hospitalized in surgical and maternity medical services in Colombia and obtained a Cronbach's alpha of 0.83 for the ENCS and a Cronbach alpha of 0.95 for the SNCS. Changing context, Torres and Buitrago (2011) used it in patients receiving oncological treatments in Colombia, with 75 patients showed a reliability of 0.91 for the experience dimension and 0.98 for the satisfaction dimension, with an alpha of Cronbach of 0.96 for the whole instrument 37.
The CUCASE in Spain version (castellan language) Cronbach's alpha was 0,8561 in the scale ENCS and 0,9744 by SNCS 0,9754 and (Gallego language) Cronbach's alpha was 0,9280 in the scale ENCS and 0,9744 by SNCS 0,9785 16.
In original scale Cronbach's alpha was 0, 91 for the experience scale and 0-96 for the satisfaction scale 6,19 and correlations between single items and the total ranged from 0-31 to 0-69 for the experience scale (24 out of 26 exceeded 0-4) and from 0 53 to 0-82 for the satisfaction scale.
The CUCASE becomes the first reliable version in Spanish in Latin America that is added to the translations of the NSNS in the world Piredda et al. (2015), studied the psychometric properties of the Italian version of the NSNS, since to date no study of construct validity had explored the factorial analysis after its development 28. He used a sample of 659 medical and surgical patients, finding differences based on the factors found by the original authors of the scale 19. The confirmatory factor analysis collected a factor for the satisfaction scale and 4 factors for the scale of the experiences: a) the lack of care or the time dedicated to the patients to satisfy their needs, b) the emotional support, related to the care that provides them with the comfort and attention they need, c) the interpersonal relationship and information, and d) the environment of trust. These results reflect the multidimensional nature of the nursing care experience.
However, in the present study, the CUCACE in the exploratory analysis by main components of the ENCS showed 7 factors and the items do not correspond to those found by Piredda 28. Neither does it happen to the SNCS in which it throws 6 factors very different from that of a factor defined in the Italian context.
The CVI in this study was 96% low to of the Turkish 18 version of the SNCS was 98%, indicating an acceptable level of content validity. In Turkish study, correlations between single items range from 0,43 to 0,89 and the internal consistency of the SNCS assessed by Cronbach's alpha is 0,96, hi in relation with our study with CVI 0,863. The means of the SNCS for the Colombian sample were consistent with previously reported English 29 and Turkish 18 and Jordan 9,25 samples. In the previous studies, items correlations ranged from .53 to .82, Cronbach's alpha was .96 and .93.
The variance explained, the number of factors and the distribution of the items in the construct analysis of the Colombian version of the SNCS substantially coincide with the Italian version 27.
Factor analysis of the scale Experiences of nursing scale (ENCS) 4 factors were extracted, that explained about 42% in Italian version 28.
In Turkey, they measured satisfaction with nursing care 18 but the measurements were not valid or reliable from the patient's perspective. Although satisfaction with care had been evaluated in Colombia, the instrument had not been validated either.
Planning care, organizing teams, providing the best nursing care in hospital care services is a priority in the light of evaluating the satisfaction of patients with care. The SNCS version (19-item scale) is adequate to measure patient satisfaction with nursing care. Our study demonstrated its usefulness in clinical settings by measuring the satisfaction of Colombian medical and surgical patients with nursing care. The results emphasize in each of the concepts that comprise the construct such as the importance of providing patients with information about their medical condition. Support family members of patients and focus more closely on the needs of patients. Both the experience and the CUCASE (Nursing Care Quality Questionnaire) should be the subject of further research in larger studies ♠