Pressure Injury Risk Assessment in Pediatric Intensive Care: A Prospective Validation Study of the Glamorgan Scale and Predictive Comparison with the Braden Q Scale

INTRODUCTION

Pressure injuries (PIs) are considered a major health problem and recognized as an indicator of patient care and safety. In addition to causing pain and increased risk of infection, PI occurrence also delays recovery, prolongs hospitalization, and increases staff workload and hospitalization costs.1–4

Pressure injury is considered a frequent adverse event in critically ill adult patients. However, the frequency of PI occurrence has increasingly been reported and studied in critically ill pediatric patients as well.5–8 The incidence and prevalence of PI in pediatric intensive care patients vary according to patient and unit characteristics. Studies have reported incidence ranging from 0.29% to 42.4%6–11 and prevalence of 2% to 28%.12–16

Because children’s skin is immature, it tends to be poorly protected, thin, and sensitive.17,18 In addition, pediatric patients are at higher risk of sepsis from changes in their skin barrier, a serious complication that increases their risk of mortality. Thus, it is essential to pay attention to maintaining their skin’s integrity.4

Pressure injury risk predictive scales have been developed specifically for the pediatric population. The Braden Q Scale (BQS) is the most widely used, although its parameters were adapted from the adult version of the scale (Braden Scale).19–21 The Glamorgan Scale (GS), which was developed in 2009 (original version in English), was specifically created for critically ill pediatric patients.22,23 In 2021, a cultural adaptation of this instrument was made into Brazilian Portuguese, named Escala de Glamorgan.24 This scale was designed to assess PI risk in children and adolescents hospitalized under intensive care from birth to 18 years of age; it is also suitable for use in preterm newborns.22,23

Cultural adaptation and validation are essential in order to offer reliable instruments to medical professionals. Instruments for assessing population health must offer accurate, valid, and interpretable data.25 Validation refers to the accuracy of results obtained from a given instrument; reliability refers to the degree of consistency of the instrument measuring the phenomenon it was intended to evaluate.26 Reliability and validity are both important properties that should be found in instruments for measurement.27–29

The use of validated instruments in medical practice provides certification for evidence-based care. Therefore, in the pursuit of improved care and PI prevention in the pediatric population, the objective of this study was to validate the GS, which was culturally adapted into Brazilian Portuguese.24 Further, the authors compared its predictive ability with that of the BQS to assess the risk of PI in pediatric ICU (PICU) patients.

METHODS

To validate the GS, the researchers developed a prospective cohort study and performed predictive comparison with the BQS.

Study Population

The sample was composed of pediatric patients admitted to the PICU of a public university hospital from February to July 2020. This unit has seven beds and admits children from 30 days up to 15 years of age. Patients were included if they were between 30 days and 14 years, 11 months, and 29 days in age; stayed at least 24 hours in the hospital; had no preexisting PI; and if their guardian agreed to their participation and signed the informed consent form.

Data Collection

Two forms were used for data collection. The first was the clinical form, which consisted of items related to demographic characteristics, clinical profile of the patients, and preventive measures adopted. In cases of PI, the form included items to characterize the injury, such as location, stage, and dressing used. The second was the form used to register the GS (Portuguese adapted version) and the BQS (Portuguese version) scores; this form was filled out by the same researcher at each assessment during the initial 24 hours of hospitalization and at each reassessment every 24 hours thereafter.20,24 The assessments were performed until discharge from the unit, hospital discharge, or death occurred.

Glamorgan and Braden Q Scales

The BQS consists of seven domains: (1) mobility, (2) activity, (3) sensory perception, (4) moisture, (5) friction and shear, (6) nutrition, and (7) tissue perfusion and oxygenation, which are scored from 1 (least favorable) to 4 (most favorable). Thus, total BQS scores range from 7 to 28 points, depending on the present impairment. At the end of the assessment, the risk is interpreted as follows: scores less than 22 indicate high risk, and scores greater than or equal to 22 indicate low risk. The lower the score, the higher the risk for developing PI.19,20

The GS is composed of nine items: (1) mobility, (2) equipment/devices/hard surface pressing or rubbing the skin, (3) significant anemia, (4) persistent fever, (5) impaired peripheral perfusion, (6) inadequate nutrition, (7) low serum albumin (<35 g/L), (8) weight below the 10th percentile, and (9) incontinence. Each item is scored based on the present impairment,22–24 and these nine scores are added to obtain a total score ranging from 0 to 42 points. The higher the score, the greater the impairment and, consequently, the higher the risk of developing PI. According to the GS criteria, there are three risk stratifications for the total score: at risk (10-14), high risk (15-19), and very high risk (20+).22–24

Validation of the GS and Predictive Comparison with the BQS

After being culturally adapted, the psychometric properties of the GS were tested to validate whether the characteristics of the original version were preserved. To this end, the researchers assessed the reliability (internal consistency) and concurrent and predictive validity. For the predictive comparison, they performed the same tests with the BQS.

Reliability

An essential aspect in assessing the quality of an instrument, especially if it is a measurement scale, is checking for its internal consistency, that is, checking if the items are correlated and measuring the same factor. The reliability of a measure refers to the degree to which the measure is free from error and therefore yields consistent results.30–32

Internal consistency

An instrument has internal consistency when its variables are all measuring the same attribute. Internal consistency can be determined with a single application of the instrument through indicators, such as Cronbach α, which verifies the homogeneity of the items so that they measure the same dimensions based on the number of scale items and the homogeneity of each item, defined as the mean of all correlations between the items. This reflects the extent to which items/subitems have similar variance and indicates the internal consistency of the measurement. The cited authors suggest a CI of 95% and a minimum level of reliability of 0.70.33–36

Concurrent and predictive validity

Concurrent validity assesses how well an instrument measures exactly what it is supposed to measure, when comparing its performance to that of another instrument that measures something similar; that is, it assesses the ability of an instrument to accurately measure the phenomenon to be studied and if it is useful for its intended purpose. An instrument can be considered valid when it can truly assess its objective (ρ > 0.7).35,37,38 In this study, concurrent validation compared the scores of the GS with those of the BQS.

The predictivity of the two scales was analyzed to determine their sensitivity in identifying PI, which was performed by analysis of the receiver operating characteristic (ROC) curve generated from the accuracy of the GS scores.

Sample Size

Given the variation in the types of instruments used, there are no absolute rules regarding a unique sample size required for the different steps of validation.39 Therefore, the sample size was estimated to represent up to 10% of PIs among the assessments, with a 2.5% standard error, β of .20, and α of .05, which result in a minimum of 554 scale applications. In addition, to detect a coefficient correlation between the scales up to 0.5 (power = 90% and α = .05), only 53 measurements are needed. To detect a sensibility and specificity up to 70% (power = 90% and α = .05), 56 measurements are needed.

Statistical Analysis

The collected data were entered into a Microsoft Excel spreadsheet to identify the absolute and relative frequencies of the values obtained by the instruments used. Descriptive statistics were performed with frequency and percentages for categorical variables, and mean, SD, median, minimum, and maximum for quantitative variables. The normality of the data was assessed using the Shapiro-Wilk test.38

The unit of the analysis was each score from the scales, measured in the visits to the PICU. The internal consistency of the instruments, which is a measurement based on the correlations among different items in the same test, was assessed using Cronbach α coefficient. When the value of Cronbach α coefficient is less than .7, the internal consistency is considered adequate. However, Cronbach α coefficient values are strongly influenced by the number of items of the measurement instrument. A small number of items per domain of an instrument can decrease the α values, thus affecting internal consistency.34,35

Classification accuracy, or predictive validity, was evaluated by means of the ROC curve and calculation of area under the curve (AUC). The AUC is a measure of ranking performance (test accuracy index), which is considered satisfactory when it has a value greater than 0.70.30–43

The correlation (equivalence) between the GS and BQS was performed using Spearman correlation coefficient (ρ). Spearman correlation coefficient is interpreted as follows: 0.00 to 0.10 signifies negligible correlation, 0.10 to 0.39 signifies weak correlation, 0.40 to 0.69 signifies moderate correlation, 0.70 to 0.89 signifies strong correlation, and 0.90 to 1.0 signifies very strong correlation.35P < .05 was considered significant.

Ethics

The project was approved by the Research Ethics Committee of Botucatu Medical School, recognized by the National Research Ethics Commission, meeting the standards of Resolution 466/12 of the National Health Council, and was approved according to Opinion No. 1,908,776.

All study participants signed two copies of the informed consent form, with one copy remaining with the research participant and the other with the researcher.

RESULTS

The sample consisted of 83 pediatric patients in the PICU (mean age, 4.02 [SD, 4.3] years), 50 of whom were male (60.2%). Of the patients, 31 (37.3%) had diagnoses related to respiratory diseases, and 55 (66.2%) used mechanical ventilation. The mean hospitalization time was 8.1 (SD, 4.7) days.

After using the scales, a total of 639 assessments were obtained, with an average of 7.7 assessments per patient. The mean total scores were 31.8 for the GS (SD, 8.2), where scores greater than 20 indicate very high risk, and 14.8 for the BQS (SD, 3.6), where scores lower than 22 indicate high risk. Table 1 shows the risk classification according to the scales’ cutoff values.

Table 1 - RISK CLASSIFICATION ACCORDING TO ASSESSMENTS (N = 83) Variable Cutoff Value n (%) Total no. measurements — 639 (100.0) Glamorgan Scale  At risk 10–14 27 (4.2)  High risk 15–19 44 (6.9)  Very high risk ≥20 568 (88.9) Braden Q Scale  Low risk ≥22 42 (6.6)  High risk <22 597 (93.4)

In the total sample (N = 83), 25 PIs developed in 18 patients, leading to an incidence of 21.6%. A total of 187 assessments were performed, with a mean of 10.3 assessments per patient and a mean time of occurrence of the outcome of 8.2 (SD, 3.5) days. The mean total scores were 36.6 (SD, 1.9) for the GS and 12.2 (SD, 2.6) for the BQS. All PIs occurred in patients who had a GS score greater than 30 and a BQS score of less than or equal to 16.

Most patients who developed PI were younger than 5 years, with a mean age of 4.9 (SD, 3.5) years and mean hospitalization time of 11.0 (SD, 3.5) days. Fasting was prescribed for all 18 patients who developed a PI (Table 2). Thirteen patients developed one PI, four patients developed two, and one patient developed four PIs. Of the 18 patients who developed PI, five (27.8%) died.

Table 2 - CHARACTERIZATION OF PATIENTS WHO DEVELOPED PRESSURE INJURIES (n = 18) Variable n (%) Sex  Female 7 (38.8)  Male 11 (61.1) Skin color  White 13 (72.2)  Brown 5 (27.7)  Black 0 (0.0) Diagnosis  Respiratory diseases 9 (38.8)  Neurological diseases 6 (33.3)  Neoplastic diseases 3 (16.6)  Heart disease 2 (11.1) Mechanical ventilation  Yes 18 (100.0)  No 0 (0.0) Drugs  Sedative 18 (100.0)  Analgesic 14 (77.7)  Vasoactive 8 (44.4)  Diuretic 5 (27.7)  Vasopressor 4 (22.2)  Anti-inflammatory 2 (11.1)  Anticonvulsant 2 (11.1)  Chemotherapeutic 2 (11.1)  Corticoid 1 (5.5)  Antibiotic 1 (61.1) Route of dietary administration  Fasting 18 (100.0)  Probea 17 (94.4)  Gastrostomy 1 (5.5) Preventive measures  Equipment changes 18 (100.0)  Decubitus changes 18 (100.0)  Pyramidal mattress 18 (100.0)  Skin hydration 18 (100.0)  Positioner cushions 18 (100.0)  Hydrocolloid 8 (44.4) Pressure injuries (n = 25)  Sacral 8 (32.0)  Occipital 7 (28.0)  Heel 7 (28.0)  Auricular pavilion 1 (4.0)  Trochanter 1 (4.0)  Knee 1 (4.0) Stage  1 13 (52.0)  2 7 (28.0)  3 3 (12.0)  Unstageable 2 (8.0) Dressing  Essential fatty acids 12 (48.0)  Hydrocolloid 6 (24.0)  Papain gel 5 (20.0)  Hydrogel 2 (8.0)

aNasogastric tube, orogastric tube, or nasoenteric tube.

Table 3 shows the subscales that form the GS and BQS. The one with the lowest individual mean score of the 187 evaluations performed by the BQS was the subscale “activity,” with a mean score of 1.0 (SD, 0.1; worst evaluation score = 1). The best mean score among the assessments was the subscale “moisture,” with a mean score of 2.7 (SD, 0.7; best assessment score = 4). For the GS, the subscale that was most associated with the development of injuries was “mobility,” with a mean score of 18.9 (SD, 2.88; worst assessment score = 20). The subscale that had the least association with PI development was “albumin,” with a mean score of 0.1 (SD, 0.2; best assessment score = 0).

Table 3 - MOST AND LEAST PREDICTIVE SUBSCALES ASSESSED FOR THE 18 PATIENTS WHO DEVELOPED PRESSURE INJURIES Pressure Injury Risk Assessment Scale Mean Median SD Glamorgan Scale  Mobility 18.9 20 2.8  Equipment 15.0 15 0.1  Anemia 0.2 0 0.4  Fever 0.2 0 0.4  Perfusion 0.4 0 0.4  Nutrition 0.8 1 0.3  Albumin 0.1 0 0.2  Weight 0.2 0 0.4  Incontinence 0.5 1 0.4 Braden Q Scale  Mobility 1.4 1 0.7  Activity 1.0 1 0.1  Sensory perception 1.6 2 0.7  Moisture 2.7 3 0.7  Friction and shear 1.4 1 0.6  Nutrition 1.7 2 0.7  Tissue perfusion and oxygenation 2.2 2 0.7

To assess internal consistency, researchers included all variables of the instruments. The BQS (Cronbach α = .7; 95% CI, 0.7–0.8) showed better internal consistency than the GS (Cronbach α = .3; 95% CI, 0.2–0.4). These results indicate that the items of the GS are less interrelated than those of the BQS.

Researchers assessed equivalence with the Spearman correlation coefficient (ρ = −0.76; P < .01) and found a strong correlation between the total scores of the GS and BQS (Figure 1). The correlation value was negative, because in the BQS, the lower the score, the higher the risk, and in the GS, the higher the score, the higher the risk.

F1Figure 1:

DIAGRAM OF DISPERSION BETWEEN THE BRADEN Q SCALE AND GLAMORGAN SCALE SCORES

Predictive validity (accuracy) was assessed by the ROC curve and AUC calculation, and values greater than 0.70 were interpreted as having satisfactory performance. Figure 2 shows good accuracy of the two scales and a similar performance between them. The GS showed an AUC value of 0.77 (CI, 0.72–0.82) versus 0.78 (CI, 0.73–0.84) for the BQS. For very high-risk scores (≥20), GS presented a sensitivity of 100.0, and a specificity of 12.3% in this sample. For BQS, the high-risk scores (<22) resulted in a sensitivity of 100.0% and a specificity of 10.7%. In addition, analysis of the ROC curve suggests lower cutoff values for GS, because a score of 17 can increase its specificity to 20%, without decreasing the 100% sensitivity.

F2Figure 2:

RECEIVER OPERATOR CHARACTERISTIC CURVE OF THE GLAMORGAN AND BRADEN Q SCALES APPLIED TO CRITICALLY ILL PATIENTS

DISCUSSION

The psychometric evaluation of the GS with a sample of 83 patients admitted to a PICU showed that the instrument is both reliable and valid in assessing the risk of PI in critically ill pediatric patients. Analysis of the mean total scores of the scales in patients who developed PI (GS mean score, 36.6; BQS mean score, 12.2) indicated that both are effective tools for predicting PI risk in the pediatric population. The patients with assessment scores indicating high risk were more likely to develop PI.

As for the profile of affected patients, their mean age of 4.9 years showed that this age group was most prone to developing PI. At this age, the skin is more susceptible to trauma because of its sensitivity.44 There was a predominance of white skin color, as in other studies,6–8 which may be related to the higher prevalence of these patients in the institution.

All 18 patients who developed PI were undergoing invasive mechanical ventilation (IMV), making IMV an important risk factor. In addition to making it difficult to maintain adequate perfusion and oxygenation, IMV also hinders patient mobilization, which may contribute to PI development.45,46 The most frequent diagnoses among study participants were respiratory diseases, corroborating the findings of other studies.6–8 Morbidities associated with the respiratory system impair oxygen supply to tissues, thus facilitating PI development.47,48

Because of the hemodynamic instability of critically ill patients, continuous-use drugs are often necessary; these can contribute to the development of PI by interfering with peripheral blood flow and tissue perfusion.49,50 The most frequently prescribed drugs were sedatives, analgesics, and vasoactive drugs. Sedative and analgesic drugs are commonly used in intensive care to reduce respiratory discomfort and improve adaptation to IMV. However, sedoanalgesia is associated with mobility deficits in patients, thus prolonging IMV times and contributing to PI development.51,52

Fasting was prescribed for the 18 patients who developed PI because of investigation and/or treatment procedures. The most used route of dietary administration was via tube, whether it be a nasogastric or orogastric tube. Inadequate food intake has a great impact on the body’s physiologic responses and is related to injury development. Situations that affect the digestive process may compromise nutrient absorption and adequate use of food.53 Thus, when enteral nutrition therapy is administered, the nutrition needs of the patient are often not fully met, thus increasing the risk of PI.

The validation of the GS was evidenced by its reliability (internal consistency and equivalence) and predictive validity. In the reliability analysis, the GS (Cronbach α = .3) showed lower internal consistency than the BQS (Cronbach α = .7). The BQS is a reflective scale, including variables that together reflect the risk of PI development. Because all the elements are correlated, when there is an increase in the risk of PI, it affects all the elements in the same way. In contrast, the GS is an additive scale, with items that increase the risk of PI, but it includes different and uncorrelated elements. Thus, by definition, the additive scale does not have a high Cronbach α.34

The Spearman correlation coefficient (ρ = −0.76; P < .01) showed a strong correlation between the total scores of the GS and BQS, corroborating the concurrent validity of the GS through comparison with another scale that has already been validated to assess the risk of PI in the pediatric population. According to the analysis of predictive validity (accuracy) as evaluated by the ROC curve and calculation of AUC, the performance of both scales was satisfactory and similar, with the GS having an AUC of 0.77 (CI, 0.72–0.82) versus 0.78 (CI, 0.73–0.84) for the BQS. In a similar study, the GS (0.91; CI, 0.87–0.94) had a better performance than the BQS (0.69; CI, 0.62–0.76).54 Therefore, the classificatory performance of the GS does not differ from that of the BQS, with both having good predictive accuracy.

Limitations

Some limitations of the present study were its monocentric nature and its inclusion of patients only from the Brazilian Unified Health System. However, these limitations do not invalidate the results aimed at validating the instrument for use in the Brazilian population. Interrater reliability was not performed because the same rater performed all evaluations; this could be a source of some variability of the scores.

CONCLUSIONS

The GS was validated for use in critically ill pediatric patients confined to the PICU. Both the GS and BQS showed similar and satisfactory accuracy, with a strong correlation coefficient. Thus, both instruments are capable of assessing the risk of PI development in critically ill pediatric patients.

REFERENCES 1. National Pressure Injury Advisory Panel (NPIAP). Pressure Injury Stages; 2016. https://cdn.ymaws.com/npiap.com/resource/resmgr/NPIAP-Staging-Poster.pdf/. Last accessed December 12, 2022. 2. Razmus I, Bergquist-Beringer S. Pressure injury prevalence and the rate of hospital- acquired pressure injury among pediatric patients in acute care. J Wound Ostomy Cont Nurs 2017;44(2):110–7. 3. Cummins KA, Watters R, Leming-Lee TS. Reducing pressure injuries in the pediatric intensive care unit. Nurs Clin North Am 2019;54(1):127–40. 4. Schlüer AB. Pressure ulcers in maturing skin—a clinical perspective. J Tissue Viability 2017;26(1):2–5. 5. Aprea V, Barón FJ, Meregalli C, Sabatini MC. Impact of a health care quality improvement intervention to prevent pressure ulcers in a pediatric intensive care unit. Arch Argent Pediatr 2018;116(4):e529–41. 6. Vocci MC, Toso LAR, Fontes CMB. Application of the Braden Q Scale at a pediatric intensive care unit. Rev Enferm UFPE Online. 2017;11(1):165–72. 7. Vocci MC, Fontes CMB, Abbade LPF. Pressure injury in the pediatric population: cohort study using the Braden Q Scale. Adv Skin Wound Care 2018;31(10):456–61. 8. Vocci MC, Onary ASS, Castro MCN, Spadotto AFO, Fontes CMB. Retrospective analysis of the application of the Braden Q Scale in pediatric intensive care. ESTIMA Braz J Enterostomal Ther 2020;18:e2820. 9. Zollo MB, Gostisha ML, Berens RJ, Schmidt JE, Weigle CG. Altered skin integrity in children admitted to a pediatric intensive care unit. J Nurs Care Qual 1996;11(2):62–7. 10. Kottner J, Wilborn D, Dassen T. Frequency of pressure ulcers in the pediatric population: a literature review and new empirical data. Int J Nurs Stud 2010;47(10):1330–40. 11. Carvalho GBC, Silva FA, Castro ME, Florêncio SF. Epidemiology and risks associated with pressure ulcer in children. Cogitare Enferm 2011;16(4):640–6. 12. Schlüer AB, Cignacco E, Müller M, Halfens RJ. The prevalence of pressure ulcers in four paediatric institutions. J Clin Nurs 2009;18(23):3244–52. 13. Crozeta K, Stocco JGD, Danski MTR, Meier MJ. Pressure ulcers in neonates and children: epidemiological and clinical profile. Rev Min Enferm 2010;14(2):233–8. 14. Bergquist-Beringer S, Gajewski B, Dunton N, Klaus S. The reliability of the national database of nursing quality indicators pressure ulcer indicator: a triangulation approach. J Nurs Care Qual 2011;26(4):292–301. 15. Chiari P, Poli M, Magli C, et al. Multicentre, prospective cohort study, to validate the Italian version of the Braden Q Scale for the risk of the pressure sores in newborns and up to 8 years old children. Assist Inferm Ric 2012;31(2):83–90. 16. Schlüer AB, Schols JM, Halfens RJ. Risk and associated factors of pressure ulcers in hospitalized children over 1 year of age. J Spec Pediatr Nurs 2014;19(1):80–9. 17. Mack MC, Chu MR, Tierney NK, et al. Water-holding and transport properties of skin stratum corneum of infants and toddlers are different from those of adults: studies in three geographical regions and four ethnic groups. Pediatr Dermatol 2016;33(3):275–82. 18. Walters RM, Khanna P, Chu M, Mack MC. Developmental changes in skin barrier and structure during the first 5 years of life. Skin Pharmacol Physiol 2016;29(3):111–8. 19. Curley MAQ, Razmus IS, Roberts KE, Wypij D. Predicting pressure ulcer risk in pediatric patients. The Braden Q Scale. Nurs Res 2004;52(1):22–31. 20. Maia ACAR, Pellegrino DMS, Blanes L, Dini GM, Ferreira LM. Translation into Portuguese and validation of the Braden Q Scale for assessing the risk of pressure ulcer in children. Rev Paul Pediatr 2011;29(3):406–14. 21. Bergstrom N, Braden B, Laguzza A. The Braden Scale for predicting pressure sore risk. Nurs Res 1987;36:205–210. 22. Willock J, Baharestani MM, Anthony D. The development of the Glamorgan Paediatric Pressure Ulcer Risk Assessment Scale. J Wound Care 2009;18(1):17–21. 23. Willock J, Baharestani M, Anthony DM. The development of the Glamorgan Paediatric Pressure Ulcer Risk Assessment Scale. J Child Young People Nurs 2007;1(5):211, 8. 24. Vocci MC, Fontes CMB, Abbade LPF. Cultural adaptation of the Glamorgan Scale to Brazilian Portuguese: pressure injury in pediatrics. Rev Latino Am Enferm 2021;29:e3424. 25. Alexandre NMC, Gallasch CH, Lima MHM, Rodrigues RCM. Reliability in the development and evaluation of measurement instruments in the health area. Rev Eletr Enferm 2013;15(3):802–9. 26. Pereira LW, Bernardi JR, Matos S, Silva CH, Goldani MZ, Bosa VL. Cross-cultural adaptation and validation of the Karitane Parenting Confidence Scale of Maternal Confidence Assessment for use in Brazil. J Pediatr (Rio J) 2018;94:192–9. 27. Cook DA, Beckman TJ. Current concepts in validity and reliability for psychometric instruments: theory and application. Am J Med 2006;119(2):166.e7–16. 28. Pittman J, Bakas T. Measurement and instrument design. J Wound Ostomy Continence Nurs 2010;37(6):603–7. 29. Salmond SS. Evaluating the reliability and validity of measurement instruments. Orthop Nurs 2008;27(1):28–30. 30. Terwee CB, Bot SD, Boer MR, et al. Quality criteria were proposed for measurement properties of health status questionnaires. J Clin Epidemiol 2007;60(1):34–42. 31. Keszei AP, Novak M, Streiner DL. Introduction to health measurement scales. J Psychosom Res 2010;68(4):319–23. 32. Polit DF, Beck CT. Fundamentos de Pesquisa em Enfermagem: Métodos, Avaliação e Utilização. 7th ed. Porto Alegre, Brazil: Artmed; 2011. 33. Bonett DG, Wright TA. Cronbach’s alpha reliability: interval estimation, hypothesis testing, and sample size planning. J Organ Behav 2015;36(1):3–15. 34. Tavakol M, Dennick R. Making sense of Cronbach’s alpha. Int J Med Educ 2011;2:53–5. 35. Miot HA. Correlation analysis in clinical and experimental studies. J Vasc Bras 2018; 17(4):275–9. 36. Heale R, Twycross A. Validity and reliability in quantitative studies. Evid Based Nurs 2015;18(3):66–7. 37. Kimberlin CL, Winterstein AG. Validity and reliability of measurement instruments used in research. Am J Health Syst Pharm 2008;65(23):2276–84. 38. Miot HA. Correlation analysis in clinical and experimental studies. J Vasc Bras 2017;16(2):88–91. 39. Miot HA. Sample size in clinical and experimental trials. J Vasc Bras 2011;10:275–8. 40. Osborne JW, Costello AB. Sample size and subject to item ratio in principal components analysis. Pract Assess Res Eval 2004;9:8. 41. Metz CE. Basic principles of ROC analysis. Semin Nucl Med 1978;8(4):283–98. 42. Polo TCF, Miot HA. Use of ROC curves in clinical and experimental studies. J Vasc Bras 2020;19:e20200186. 43. Hoo ZH, Candlish J, Teare D. What is an ROC curve? Emerg Med J 2017;34(6):357–9. 44. Cestari SCP. Pediatric Dermatology. 1st ed. São Paulo, Brazil: Editora Atheneu; 2012. 45. Loudet CI, Marchena MC, Maradeo MR, et al. Reducing pressure ulcers in patients with prolonged acute mechanical ventilation: a quasiexperimental study. Rev Bras Ter Intensiva 2017;29(1):39–46. 46. Miske LJ, Stetzer M, Garcia M, Stellar JJ. Airways and injuries: protecting our pediatric patients from respiratory device-related pressure injuries. Crit Care Nurs Clin North Am 2017;29(2):187–204. 47. Curley MAQ, Hasbani NR, Quigley SM, et al. Predicting pressure injury risk in pediatric patients: the Braden QD Scale. J Pediatr 2018;192(2):189–95. 48. Smith HA, Moore Z, Tan MH. Cohort study to determine the risk of pressure ulcers and developing a care bundle within a paediatric intensive care unit setting. Intensive Crit Care Nurs 2019;53:68–72. 49. Cox J, Roche S. Vasopressors and development of pressure ulcers in adult critical care patients. Am J Crit Care 2015;24(6):501–10. 50. Tallo FS, Vendrame LS, Lopes RD, Lopes AC. Invasive mechanical ventilation in the emergency room: a review for clinicians. Rev Bras Clin Med 2013;11(1):48–54. 51. Santos LRCL, Lino AIA. Risks of pressure injury: application of the Braden Scale in intensive care. ESTIMA Braz J Enterostomal Ther 2018;16:e0818. 52. Barbosa TP, Beccaria LM, Silva DC, Bastos AS. Association between sedation and adverse events in intensive care patients. Acta Paul Enferm 2018,31(2):194–200. 53. Simony RF, Chaud DMA, Abreu ES, Assis SMB. Nutritional status of neurological patients with reduced mobility. J Hum Growth Dev 2014;24(1):42–8. 54. Anthony D, Willock J, Baharestani M. A comparison of Braden Q, Garvin and Glamorgan risk assessment scales in paediatrics. J Tissue Viability 2010;19(3):98–105.

留言 (0)

沒有登入
gif