Incidence of pressure ulcers in cardiopulmonary intensive care unit patients *

Objective: Identify and analyze the incidence coefficients of pressure ulcers (PU) and the risk factors for PU development in critical patients with cardiopulmonary diseases. Method: A prospective cohort study conducted in the cardiopulmonary intensive care unit (ICU) of a large hospital in the city of São Paulo, from November 2013 to February 2014. In total, 370 patients over 18 years old who did not present PU at admission and who had been in the ICU for less than 24 hours were studied. Data were analyzed by univariate and multivariate analyses (Classification And Regression Tree CART). Results: The incidence coefficients of PU were: 11.0% for total participants, 8.0% for male and 3.0% for female patients (p=0.018); 10.0% for white patients and 6.5% for patients 60 years or older. The main risk factors were length of stay in the ICU for 9.5 days or more, age 42.5 years or older, and being of the white race. Conclusion: This study was related to the epidemiology of PU in critical patients with cardiopulmonary diseases, facilitating the planning of specific preventive care for these patients. DeScriPtOrS Pressure Ulcer; Incidence; Intensive Care Units; Inpatients; Nursing Care; Epidemiologic Studies. Correspondence Addressed to: Ticiane Carolina Gonçalves Faustino Campanili Rua Tagipuru, 1060, Apto. 83, A1 – Barra Funda CEP 01156-000 – São Paulo, SP, Brasil ticiane.faustino@gmail.com USP


INTRODUCTION
Skin injuries are a great challenge to health professionals who provide care to critical patients, considering that, in clinical practice, the occurrence of pressure ulcers (PU) have become more frequent, with an increase in treatment costs, longer hospitalization periods, discomfort and negative impact on the quality of service provided and the quality of life of patients (1)(2)(3) .
Skin care programs for the prevention of injuries, based on international guidelines (1,4) , have contributed to a reduced number of skin injuries in intensive care units (ICUs), although not all risk factors for these injuries are known, due to the specificities of each patient, each associated disease (5) and each scenario.Particularly in cardio pulmonology, due to improved diagnostic and therapeutic methods, repair and corrective surgeries have been indicated late in the process or for patients with serious conditions.This has caused a greater number of risky situations, as these procedures are compromised by comorbidities and other pre-existing diseases (6) .The real dimensions of skin injury problems in cardiopulmonary ICU patients has raised the interest of professionals who provide care to these patients.
PU is among skin injuries frequently found in critical patients, highlighted due to its multifactorial etiology (1,7) .Several factors are associated with the development of PU, including, specifically for critical patients, extrinsic compression associated with advanced age, nutritional deficiency, humidity, immobility and bed rest, reduced tissue perfusion, use of vasoactive drugs, sedation, and comorbidities like diabetes mellitus and vascular disease (8)(9) .
Reduced sensory perception caused by anesthetic agents and longer periods of immobility due to the impossibility of patient mobilization during the operation (8) facilitate the development of tissue hypoperfusion areas after cardiopulmonary surgeries.Patients whose sensory perception is impaired, that is, who are not able to detect sensations that indicate the need to change position, are among the most susceptible to PU development (6,(8)(9) .
In surgery patients, PU usually appears in the period that includes the surgery up to five days after the exposure to the compression applied in the surgery room, although here it is often unnoticed or described as hyperemia only, (8)(9) .The cardiovascular system tolerance to a surgery procedure is influenced by pharmacological agents that keep venous return and proper systemic vascular resistance.However, the circulatory and respiratory systems may be affected during the surgery, causing hypotension and consequent alteration to tissue perfusion, requiring vasoactive drugs to improve blood circulation and preserve cerebral and myocardial tissue irrigation (5,(8)(9) .
In normal conditions, the capillary pressure in skin artery termination is around 32 mmHg, while in venous artery termination, it is about 12 mmHg.Some situations may affect skin perfusion during the patient period in an ICU (1,(8)(9) , such as: reduced blood pressure due to cardiovascular alteration, systemic inflammatory response syndrome (SIRS), septic shock, hemorrhagic shock and shock due to drug use; external pressure kept over 32 mmHg for an extended period on rigid tables or hospital beds without a proper support surface; hemodynamic stability with elevated external pressure under the body or hemodynamic stability with continuous external pressure.In any of these situations, critical patients have interrupted blood flow to the area under pressure, affecting tissue oxygenation and nutrition in that area, with the possibility of ischemia, hypoxia, tissue acidosis, edema and tissue necrosis (8) .
Tissue tolerance to pressure and ischemia depends on tissue nature and it is influenced by the ability of the skin and supporting structures to redistribute the pressure applied (6,(8)(9) .Critical determinants for the development of PU have been widely discussed by experts for over 30 years and, every day, new factors are discovered regarding the specificity of each patient and each disease (9) .
Considering that health institutes seek to improve the service quality and patient safety, since the 1980s in Brazil, the occurrence of new cases of PU have been used as indicators of health service quality, especially in nursing, requiring the evaluation of results from prevention protocol implementation processes.More recently, in April 2013, the Ministry of Health created the National Patient Safety Program (PNSP), through Ministerial Directive MS/GM 529/2013, and one of the objectives of this program is to monitor PU incidence to minimize it in hospitalized patients (10) .
In Brazil, the incidence of PU in critical patients has been explored in few units, especially in the state of São Paulo.National studies show a great variation either in PU incidence coefficients found in critical patients (10.0% to 62.5%) (11)(12) or in the evaluation periods (1 to 15 months) (11,12) .International studies show PU incidence in ICU between 3.2% and 39.0% (13)(14)(15)(16)(17) .
Considering that critical patients are highly vulnerable to the development of skin injuries and that these injuries involve specific risk factors, depending on the underlying disease and associated conditions, this study was conducted to identify and analyze the incidence coefficients of PU and the risk factors for PU development in cardiopulmonary intensive care unit patients.

METHOD
This is an epidemiological prospective cohort study conducted in the cardiopulmonary ICU of a large public hospital in the city of São Paulo.The study sample had 370 patients admitted to the ICU from November 26, 2013 to February 26, 2014, who met the following inclusion criteria: the patients had to be at least 18 years old, have no PU at ICU admission, be in the ICU for less than 24 hours, and present ≤18 Braden scale.
Data were collected about the following demographic and clinical variables: age, ethnicity, gender, weight, height, body mass index (BMI), comorbidities and prior habits (diabetes mellitus -DM, systemic arterial hypertension -SAH, dyslipidemia -DL, encephalic vascular accident -EVA, atrial fibrillation -AF and cigarette smoking, type of surgical procedure, time of procedure, use and time of extracorporeal circulation -ECC, use of blood products, use and concentration of vasoactive drugs at admission, type and number of artifacts (venous and arterial catheters, drains), hospitalization period, date of PU onset, and ICU discharge or death.Besides this instrument, this study also used: Braden scale, Glasgow Coma Scale, Ramsay scale and Acute Physiology and Chronic Health Evolution (APACHE II).
Sociodemographic data, comorbidities, habits and surgery data were collected at admission, while the other clinical data, were collected on a daily basis during the ICU period.
The results were expressed in mean and standard deviation values for continuous variables, and absolute and relative frequencies for categorical variables.The associations of PU incidence with demographic, clinical and surgical variables used Fisher exact test for categorical variables, the Student's test for numerical variables that presented normal distribution and Wilcoxon-Mann-Whitney test for variables with other distribution.After univariate analyses, a decision tree was developed with the Classification And Regression Tree (CART) algorithm to identify risk factors for PU development.This statistical analysis technique was selected for its easy result interpretation through charts (tree).Another advantage of the CART is that it does not depend on hypothesis tests.Due to a low prevalence of the studied effects, logistic regression, frequently used in cohort studies, could lead to estimation problems in this study.The analyses were conducted using R 3.1.0,with 5.0% significance level.
The study project was approved by the Research Ethics Committees of the institution and hospital (process nº 20780713.4.0000.5392)and followed the ethical standards.
In total, 40 PUs were detected in 40 patients, resulting in a global incidence coefficient of 11.0%: 8.0% in male patients and 3.0% in female patients (p=0.018); and 10.0% in white patients.Regarding age, 6.5% incidence was observed in people 60 years or older (Table 1).
Also with regard to age, a statistically significant difference (p=0.014) was observed between the mean values for patients with PU (63 years; SD=13.4) and without PU (56.5 years; SD=16.0).
Of 40 PUs found in this study, 16 (40.0%)were classified as category I, 16 (40.0%)as category II, and 8 (20.0%) as suspected deep tissue injury, affecting the sacral region.The PUs were identified from day 2 to day 23 of the ICU period; with day 3 being the day with the greatest number of PUs diagnosed (eight injuries in the sacral region and two in the calcaneus. Among the clinical quantitative variables, the groups were significantly different, except for ICU period, longer for patients with PU (14.0;SD=12.0)(p<0.001) and, among the surgical variables, longer ECC time in the group of patients with PU (120.0;SD=68.2) (p=0.030)(Table 2).
Variables DM and use of supporting surface presented a statistically significant difference between the groups (p=0.004;p<0.001, respectively) with and without PU (Table 3).
According to the CART analysis, three factors were observed that better identified the group with PU, that is, the risk factors for the development of PU: ICU period of 9.5 days or more, age of 42.5 years or older and being of white race (Figure 1).

DISCUSSION
In the last decade, international studies in ICU have shown PU incidence coefficients of PU between 3.2% and 39.0% (13)(14)(15)(16)(17) .In post-op cardiac surgery patients (18) , an incidence of 16.4% was found.In a retrospective study (19) with 2,695 adult surgical patients, conducted in an American quaternary general hospital, an incidence of 11.0% was found, confirming the findings of this study.
In Brazil, the incidence found in ICUs ranges from 10.0 (11) to 62.5% ( 12) .In a study conducted in four distinct ICUs (two neurological units, one cardiac unit and one general unit) of two private hospitals in the city of São Paulo, the authors obtained a practically identical result (11.1%) (20) to the results from this study.The lowest index found in Brazil (10, 0%) -also similar to this study -was identified in a neurosurgical ICU (11) , while in another study (12) , conducted in an ICU of a large tertiary hospital located in the state of São Paulo, the highest incidence was 62.5%.
In a study (21) about the incidence of PU in a university hospital that used the same methodology as this study, the authors found 41.0% coefficient among critical patients before the implementation of a prevention protocol.More recently, in 2012, one of the same authors found 23.1% incidence after the implementation of a prevention protocol, which is still very high, despite a significant reduction (4) .
Triggering or risk factors are associated with the development of PU (22)(23)(24) .In a retrospective study using 222 medical charts of patients with different diseases who were submitted to surgeries lasting more than two hours, the authors related the development of PU to factors that affected tissue perfusion, like hypertension and diabetes (25) .In a comprehensive review of the literature on the use of vasopressors in critical patients and the development of PU (26) , the use of a vasopressor (noradrenaline) along with age and extended stay in the ICU were found to be risk factors for the development of PU.In another study (19) , the risk factor was the use of blood products used during the surgery.Other publications (26)(27)(28) highlight advanced age, diabetes mellitus and arterial hypertension, including body weight and chronic diseases (vasculopathies, neuropathies and anemia) as risk factors.Two of the factors highlighted (ICU period and age) were confirmed in this study.
In a recent systematic review (54 studies) conducted to identify independent risk factors for PU (27) , the authors concluded that there are primary and secondary factors involved in the development of PU: mobility/activity, perfu-  Today, nursing professionals are more and more encouraged to identify risk factors early and explore them scientifically.The literature shows that more risk factors have to be recognized and analyzed in specific populations, so that care plans can be better developed (20)(21)(22)(23) .
The risk factors found in this study are also confirmed partly by the results of other studies, which identified age above 69 years, ICU period above eight days (28) ; being of the white race as risk factors.
Age is a widely discussed factor, alone or combined with other variables like cardiovascular diseases, lifestyle and ICU period, as it involves physiological alterations related to the aging process.Loss of muscle mass, reduced inflammatory response, reduced albumin levels and subcutaneous tissue associated with reduced dermal-epidermal junction cause a higher risk of skin integrity rupture, as the skin does not have the proper ability to redistribute the mechanical load applied, presenting an important item for risk classification (28)(29)(30)(31) .The Braden scale, although commonly used and studied, at national and international levels, does not include age (1) .Experts claim risk scales should be developed according to the specific risk factors of each population, based on multivariate analyses.In clinical practice, risk assessment scales that include age as one of the factors involved in PU development should be used, such as the Cubin-Jackson scale (1) .
Regarding the ICU period, in a prospective study with 140 patients conducted in Turkey, the time spent by the patients in ICU was found to be a risk factor for the development of PU (p<0.001),always associated with mobility and level of consciousness of patients (31) , which is similar to the findings of this study.In this study, the patients were confined to bed and had reduced mobility due to the great number of artifacts inserted in the skin, use of sedatives and the patient's fear of feeling pain and discomfort when moving.
A longer period in the cardiopulmonary ICU is necessitated due to to the recovery of the physiological condition of patients submitted to invasive procedures, and the use of technologies that allow organ reconditioning and control of associated diseases.The population with cardiopulmonary alterations seeks health services late due to alternative clinical treatments, making surgeries the last option.The patients who developed PU in this study presented impaired skin integrity, predominantly on the third day of hospitalization, confirming the findings of another study (29) that describes onset of 98.0% of PUs in the first three days of post-operative hospitalization.In two other studies conducted with surgical (30) and critical (20) patients, the authors found out that PUs were diagnosed on the first two days of ICU hospitalization and that preventive measures should be used during and after surgical procedures.In these studies, the sacral region was also the most affected area, with category II injuries, confirming the results of this study and those found in the literature.The literature shows category II PUs are more frequent in patients with systemic alterations that affect tissue perfusion, such as cardiorespiratory diseases and diabetes mellitus, while category I PUs are found especially in patients in the immediate post-operative period (1,14,28) .The preferential region is explained by the dorsal decubitus patient position for long periods, hemodynamic instability and use of artifacts that do not allow or do not favor a change in position in the first 24 hours after surgery in the ICU.Also regarding cardiac surgery, immediate sternal instability does not encourage nursing staff to early patient mobilization.These patients have their sternum sutured with steel threads that allow more effective movements on a bed only after extubation, which happens, on average, six to 24 hours after ICU admission, when the patient can make more active movements, get up and sit on an armchair (1,28) .
In two studies (28,30) , race was an independent predictor of PU development; however, these findings were contradictory, as in one of these studies (28) , white race was associated with increased risk for PU (as in this study) and, in the other study (30) , black race was similarly associated.The authors of both studies conclude that race and gender are not significant independent factors for the development of PU, but they are part of a complex interaction of triggering factors for this injury.
In clinical practice, an association of intrinsic and extrinsic factors for the development of PU in critical patients is a real fact, and they may be avoided by the health team's ability of early identification of these predisposing risk factors or factors associated with the development of this type of injury.
The quality of health service has been widely discussed at national and international levels, due to the high cost of service maintenance, scarcity of available resources and the global aging of the population.It should be noted that the quality of service provided has also been evaluated with re-gard to skin injury, identifying the services that prevent it, and not the services that best treat it.By recognizing that successful prevention of PU is largely dependent on the knowledge and skills of health professionals, it is mandatory to understand individual and institutional factors that influence this knowledge and the subsequent use of scientific evidence.Systematized strategies of PU prevention may be planned and used in institutions, to increasingly reduce and improve the incidence coefficients in specific populations (1,28) and in critical post-op cardiopulmonary patients.

CONCLUSION
The results of this study showed the global PU incidence coefficient was 11.0%, with predominance of male patients, white race and patients 60 years or older.When analyzing the risk factors or predictors of PU development, the results were; period in the ICU of 9.5 days or more, age of 42.5 years or older, and being of the white race.