Characteristics and associated risk factors of exposure keratopathy among ventilated patients in intensive care units in Jordan
Article information
Abstract
Background
Exposure keratopathy is the most common ocular surface disorder in ventilated patients due to poor eyelid closure, decreased blink reflex, and the inability to produce tears. Healthcare providers in intensive care units (ICUs) play a significant role in preventing exposure keratopathy through appropriate eyelid taping and eye ointments.
Methods
This is a cross-sectional study to describe the characteristics and factors associated with exposure keratopathy in all mechanically ventilated patients admitted to an adult ICU between February and June 2023. Patients were examined for corneal changes using a corneal fluorescein staining test with a cobalt blue filter indirect ophthalmoscope.
Results
Of 156 ventilated patients included in this study, 42.3% had exposure keratopathy, 13.5% had lagophthalmos, and 26.9% of patients had chemosis. For patients with a Glasgow Coma Scale (GCS) score of 3, the odds ratio of exposure keratopathy was 21.47 (95% CI, 2.82–163.05). The use of inotropes increased the odds ratio to 35.55 (95% CI, 3.41–369.90), whereas a hospital stay >7.23 days increased the odds ratio to 43.59 (95% CI, 15.66–1,316.32).
Conclusions
The frequency of exposure keratopathy is high and is underestimated in ventilated patients, with lower GCS and increased hospital length of stay as the main risk factors. Prioritizing eye care in ventilated patients with low GCS scores or prolonged ICU stays is essential to reduce exposure keratopathy.
INTRODUCTION
Exposure keratopathy (EK) is corneal damage that mainly develops from prolonged exposure of the eye surface to the outside environment [1]. EK is the most common ocular problem among ventilated patients in intensive care units (ICUs) [2]. As a clinical disorder, EK is characterized by poor or incomplete eyelid closure and impaired tear production, leading to different types of corneal injuries [3]. During hospitalization, medical staff prioritize essential vital functions, managing life-threatening illnesses, treating organ failures, and stabilizing the medical status of the patient, with eye care becoming a secondary concern [4]. ICU patients can experience several ocular surface disorders (OSDs), ranging from minor ocular complications such as corneal dryness to more severe conditions such as corneal perforation and blindness [5].
ICU patients experience reduced nonfunctional tear production because of an imbalance between fluid and electrolyte, decreased tear immunity, and impaired blink reflexes because of the administration of muscle relaxants and sedative agents [6]. The impairment of all these mechanisms in ICU patients considerably increases their risk of OSD [7]. This may be exacerbated by decreased consciousness produced by many pathological body processes or induced by medications such as sedatives and muscle relaxants [8]. The risk of such ocular conditions is highly increased with invasive ventilation because of the length of stay in the ICU and the numerous accompanying nursing procedures, sedative agents, and effects of positive pressure ventilation [8]. A study showed a 54.3% prevalence of EK in ventilated patients compared to that of 5.1% in patients with non-invasive ventilation or no ventilation support [9]. A review study reported the occurrence of EK in about 3.6%–60% of ventilated patients in ICUs, with the highest prevalence observed on days 2 to 7 of the hospital stay [10]. Ocular surface complications such as conjunctivitis, corneal abrasion, and corneal ulcers range from 42% to 60% in sedated patients [11]. The incidence of eye disorders in critically ill patients occurred an average of 6 days after admission to the ICU [12]. Such patients have increased risk of dry eye, superficial corneal abrasions, conjunctivitis, keratitis, corneal ulcers, and corneal perforations that can lead to vision loss [7]. Conversely, a donated cornea might be rejected because of corneal dryness, which affects the number of successful corneal grafts in patients with ocular disorders.
Standardized preventive measures for patients in an ICU are crucial to reducing EK. To illustrate, standardized protocols require healthcare providers to maintain patient eye hydration using artificial tears or lubricants. Adherence to this protocol is vital for reducing and minimizing EK [13]. Various techniques prevent eye injuries, including normal saline irrigation, eye drops, tape, paraffin-based gauze, ointments, gels, and polyethylene. The use of polyethylene is the usual preventative treatment for unconscious patients [14]. Furthermore, the use of moisture chambers is associated with more effective corneal protection than lubrication [15].
Eye care education is highly beneficial in reducing the incidence of EK. Educating ICU staff to prevent EK is an easy, cheap, and effective method of eye care. Finally, the primary purpose of managing EK should be prevention rather than early diagnosis [16]. Neglected eye care in ICUs is a significant concern that is often underappreciated, especially in developing countries where the nurse-to-patient ratio is not optimal and does not fit the universal protocol [1]. Lack of clear eye protocol is another problem that affects the eye health of patients. A simple eye care protocol with simple practices can prevent significant problems among sedated and ventilated patients.
Based on a review of the literature, only one study was conducted in Jordan to assess the development of EK in the ICU in one teaching hospital [17]. In this study, we intend to update and add other local characteristics of EK among ventilated patients in different health sectors. This data will add to international results about the development of EK. This observational cross-sectional study describes the characteristics and risk factors associated with development of EK among ventilated patients in ICUs. It assessed patients after 24 hours of mechanical ventilation and determined the risk factors related to the occurrence of EK. We aim to promote the concept of eye care and recommend an eye care protocol to the hospital agency because no prevention protocols are implanted in our hospitals.
MATERIALS AND METHODS
Ethical Consideration
Ethical approval for this study was obtained from the Ethics Committee of Scientific Research at the University of Jordan (No. 1020234517). The participants’ responsible family members were asked to sign an informed consent after explaining the study and the rights of the participants. The informed consent was printed and contained a description of the study purpose, guaranteed confidentiality and anonymity, the right to withdraw from the study at any time without any consequences, and the rights of the participants. Although the hospitals involved in this study did not have established protocols or preventive measures for EK, this study addressed this gap and attempted to minimize potential risks. The research team implemented a series of proactive interventions once they discovered the patients with EK using the following steps: (1) reported it to the nurses; (2) washed the eyes thoroughly after fluorescein staining; (3) administered lubricant to each patient at risk after the procedure was finished; (4) held educational sessions for the nurses to establish eye care; and (5) recommended eye protocols based on the results. This reflects the commitment of the research team to adhere to the highest ethical standards, prioritizing the safety and well-being of the participants throughout the research process.
Design, Setting, and Sample
This study adopted a quantitative descriptive cross-sectional design to describe the characteristics and factors associated with EK among adult ventilated patients in ICUs. The study was conducted at the adult ICUs of two university teaching hospitals and two private medical centers in Jordan between February and June 2023. The first university teaching hospital had 40 beds, the second had 20 beds, and the private hospital had a capacity of 22 beds. This study was conducted in hospitals accredited by the Joint Commission International (JCI). The JCI is an independent, non-profit organization that credits and certifies healthcare organizations and programs globally [18].
This study recruited participants from private and public hospitals. All selected health settings adhere to standards set by the JCI accreditation body. Hence, our sample is relatively homogeneous regarding care practice, although the ICUs in the participating hospitals did not have a standardized protocol. No prophylactic measures such as eye-drops, ocular lubrication, taping, or other preventive procedures were used for patients, which in turn reduced the effect of confounding factors. All adult patients were admitted to the ICU for at least 24 hours for mechanical ventilation, and spontaneous blink reflexes were not included (Figure 1). The total number included in this study was 156 adults. The required sample was calculated using G*Power 3.1 (Heinrich-Heine-Universität Düsseldorf) [19] with an alpha of 0.05, a power of 0.80, and a small to moderate effect size of 0.40, yielding a minimum required sample size of 135 subjects. To be conservative, we added 10% to compensate for potential missing or incomplete data. Thus, the total proposed sample size was 150 subjects. Patients with recently or previously known ocular disorders, such as acute ocular injuries or orbital trauma, and any patient with a recent admission (less than 24 hours) were excluded.
Flowchart of patient selection and inclusion criteria. ICU: intensive care unit.
Measures
The dependent variable of this study was the presence or absence of EK. The independent variables were the study-related demographic and clinical variables, which will be discussed in the following paragraph. The researcher used the corneal fluorescein staining test to assess corneal changes, the most appropriate tool for detecting OSD in routine clinical practice [20]. Both of the patient's eyes were stained and evaluated using an indirect ophthalmoscope with a cobalt blue filter. To grade the severity of EK, the Mexican Dry Eye Disease Expert Panel proposes a practical and straightforward scale that allows classification and grading of dry eye diseases based on severity [21]. This scale is reintroduced from the National Eye Institute scale, which divides the cornea into five main quadrants. Each affected area is assigned 1–3 points depending on the distribution and amount of staining [22]. Instead of counting dots in each affected area, which is difficult and time-consuming in some cases, the Mexican panel classifies the severity of dryness based on corneal quadrant staining pattern involvement. The severity classification is as follows: mild, affecting only one peripheral quadrant of the cornea; moderate, two quadrants involved in staining or partial staining of the center part of the cornea; and severe, complete central staining of the cornea or more than three quadrants exhibiting severe epithelial damage. The consideration of complete and even partial staining of the central cornea is a moderate and severe case because it is associated with more optical disturbances affecting visual performance [23]. Additionally, the researchers developed a coding manual including demographic data, clinical data, and contextual factors based on the literature.
Demographic data were age and gender. Clinical data comprised medical diagnoses, length of hospital stay (in days), Glasgow Coma Scale (GCS) score, history of hypertension, history of diabetes mellitus (DM), use of sedative drugs, use of muscle relaxants, use of cardiac medications including inotropes, eye taping, presence of chemosis, presence of lagophthalmos, arterial blood gas readings (pH, PaCO2, PaO2, HCO3, SaO2), hemoglobin, hematocrit, white blood cell count, platelet count, potassium, sodium, creatinine, and C-reactive protein. Contextual factors included the nurse-patient ratio and the type of nursing schedule shift.
Statistical Analysis
Statistical analysis was performed using SPSS software version 26.0 (IBM Corp.). Descriptive statistics are reported as mean (standard deviation [SD]) for continuous variables and frequencies for categorical factors. Additionally, non-parametric tests, including the Mann-Whitney and chi-square tests, were used to identify factors associated with EK occurrence. Regression analysis was carried out to detect the odds ratios (ORs) for significant covariates and factors related to EK.
RESULTS
Of 156 adult sedated and/or ventilated patients in ICUs, 42.3% (n=66) had EK. Regarding the ophthalmic-related factors, 13.5% (n=21) had lagophthalmos, 26.9% (n=42) had chemosis (n=42), while only 0.6% (n=1) underwent eye taping. In patients with EK (n=66), the condition was mild in 17.9% (n=28), moderate in 14.1% (n=22), and severe in 10.3% (n=16). The site of the EK was on the right eye in 7.1% (n=11), on the left eye in 11.5% (n=18), and on both eyes in 23.7% (n=37) (Table 1). The ages of the patients ranged from 18 to 86 years, with a mean age of 58.74 (SD, 16.26). In addition, 44.2% were female (n=69) and 55.8% (n=87) were male. Of the screened patients, 73.1% (n=114) were from university teaching hospitals and 26.9% (n=42) were from private hospitals. The mean length of hospital stay was 7.23 days (SD, 4.99; range, 2–30 days) (Table 2).
Prevalence, characteristic, grade, and site of EK (n=156)
Demographic variables among the study sample (n=156)
Regarding the clinical characteristics of the included patients, 6.4% were diagnosed with renal disorders (n=10), 21.8% had respiratory issues (n=34), 14.7% had cardiac conditions (n=23), 34% had brain disorders (n=53), 10.9% had cancer-related disorders (n=17), 3.8% had gastrointestinal issues (n= 6), 1.9% had liver diseases (n=3), and 6.4% had sepsis (n=10). The results showed that the hospital length of stay ranged from 2 to 30 days with a mean of 7.23 (SD, 4.99). Hypertension occurred in 71.8% (n=112) of patients, and 39.7% (n=62) had type 2 DM. The GCS scores among the patients were as follows: 34.6% (n=45) had a score of 3 of 15, 5.1% (n=8) had a score of 4, 30.8% (n=48) had a score of 5, and 29.5% (n=46) had a score of 6. Approximately 76.3% of patients were administered sedatives (n=119), only 9% of patients received muscle relaxants (n=14), and a percentage of the patients used cardiac medications. Additionally, 30.8% of the patients used inotropes (n=48) (Table 3).
Description of clinical variables among ventilated patients (n=156)
This study revealed that chemosis and lagophthalmos had a significant relationship with the development of EK (P˂0.001). Also, hypertension (P<0.001) and type 2 DM (P<0.001) were comorbidities associated with the development of EK. However, there was no significant association between EK and medical diagnosis. The hospital length of stay and GCS score also were associated with the development of EK (P˂0.001). Regarding certain medications given in ICUs, the use of sedatives (P<0.001), muscle relaxants (P<0.001), cardiac medications (P=0.012), and inotropes (P<0.001) was significant among patients with EK. The level of HCO3 showed a significant relationship (P=0.017). Moreover, the results showed a significant relationship between the nurse-to-patient ratio in ICUs and the development of EK (P=0.022). The standard nurse-to-patient ratio in the ICU to deliver optimal care is 1:1, anything greater may affect patient safety (Table 4) [24].
Significant factors associated with the development of EK (n=156)
The final predictive model of EK risk consisted of patients with a GCS score of 3, using inotropes, and with an average hospital stay >7.23 days (Table 5). The mean hospital stay of 7.23 days was used to categorize the patients, with values of 0–7.23 assigned to category 1 and values >7.23 assigned to category 2. This dichotomization was used to facilitate the analysis. Hypertension, type 2 DM, sedatives, muscle relaxants, cardiac medications, inotropes, chemosis, and lagophthalmos were included in the predictive model but were not significant. The OR values were adjusted using regression analysis to account for potential confounders, including age and gender. Further analysis was performed using receiver operating characteristics to detect the prediction potential for ophthalmic-related factors in developing EK. The prediction potentials for chemosis and lagophthalmos were 73.9% and 62.0%, respectively (Figure 2).
Predictors of EK development in the sample (n=156)
Receiver operating characteristic (ROC) curve for lagophthalmos and chemosis.
DISCUSSION
In this study, EK occurred in 42.3% of patients, in agreement with previous studies. A retrospective cross-sectional study on the prevalence of EK conducted in Urmia Imam Khomeini Hospital in Iran reported EK in 33.4% of patients [25]. Another retrospective cross-sectional study was conducted in the USA to assess the frequency and severity of EK in critically ill patients. It found that 112 of the 205 patients (54.6%) had some degree of EK, and 58 (51.7%) had severe EK [26]. Moreover, a prospective study conducted in Jordan reported the occurrence of EK in 57% of 74 patients recruited from one teaching hospital where the eye care protocol was not followed [17] (Table 6). Ocular surface complications such as conjunctivitis, corneal abrasion, and corneal ulcers range from 42% to 60% in sedated patients [11].
Summary of previous studies on exposure keratopathy
Of comorbidities, hypertension was significantly associated with EK. Systematic hypertension decreases the production of tear film, which is a risk factor for EK [27]. Ocular surface abnormalities are common in patients with DM, who are more likely to develop dry eye disease, corneal epithelial fragility, decreased corneal sensitivity, abnormal wound healing, and infected corneal ulceration. Furthermore, around 47%–64% of patients with DM had experienced previous keratopathy [28]. The prediction model revealed that patients with a length of stay longer than seven days are at higher risk of EK. Most ocular complications are seen from days 5 to 7, and approximately 75% of patients develop corneal injuries between days 7 and 10 of admission [29,30].
EK was also associated with sedative drugs and muscle relaxants. Patients on mechanical ventilation who underwent sedation have decreased physiological function of the eye and eye muscles, which alters blinking and the tear film [31]. Reduced muscle tone contributes to lagophthalmos in ventilated patients and also can lead to corneal exposure [32]. Mechanically ventilated patients experience increased jugular venous pressure, which results in fluid retention, chemosis, and conjunctival edema that leads to a limitation in lid closure [3]. The level of HCO3 had a significant association with EK occurrence, and patients with lower levels of HCO3 had a higher risk of EK. Patients who had metabolic derangements and multiple organ dysfunction had a higher risk of impaired ocular protective mechanisms in critical care units [33]. Metabolic disorders and fluid electrolyte imbalances were predictive risk factors for EK [34,35]. A plausible explanation is that kidneys are the primary regulators of acid-base homeostasis in the human body. To prevent HCO3 loss, filtered HCO3 should be reabsorbed mainly via proximal hydrogen (H⁺) secretion. The electroneutral Na⁺/H⁺ exchanger 3 (NHE3) initiates this process in the luminal membrane with a specific kidney-type electrogenic Na⁺/HCO3 cotransporter. Theoretically, a defect in NHE3 or NBC1 can reduce the reabsorption of filtered HCO3. Patients with inactivate mutations in NBC1 can experience ocular abnormalities such as EK, glaucoma, and cataracts [35]. While our study demonstrated an association between the level of HCO3 and EK occurrence, further research is needed to understand this link.
This is the first study in Jordan to investigate the nurse-patient ratio in ICUs with EK occurrence. ICU nurses focus on life-threatening problems and not on prevention of eye problems unless the ocular complications become visible [36]. In such cases, the risk of eye disorders increases and leads to impairment during recovery [37]. Moreover, early detection of EK is the most crucial step in nursing care. Assessment of eyes should be routine in the ICU as a part of the nursing care of ventilated patients. Consequently, ICU healthcare providers should have the knowledge and skills required for optimum care and management of EK. This includes training for eye care and assessment of corneal damage.
The main inconsistency in caring for critically ill patients, who may have several ocular disorders, is the lack of a standardized eye care protocol. Using a simple eye care protocol can prevent EK and is simple in routine clinical practice. The recommended eye care protocol depends on the condition of the patients, the eye status determined every 8 hours, and proper eye lubrication and taping. In sedated and ventilated patients, lubrication was applied every 6 hours, and lubrication was used only at night if the patient was not sedated [3].
Based on the study findings, an eye care policy should be developed based on a risk scale assessment for all patients admitted to ICUs as part of the daily assessment and follow-up routine to promote early detection of EK. This helps to manage and treat ocular complications and to prevent visual disturbances or loss. The risk assessment scale should follow a predictor model and algorithm to detect high-risk patients through close observation, give more attention to early assessment, and report any changes to the ocular surface. Maintaining a healthy ocular surface also supports corneal donation, helping thousands of patients.
Although this study provides empirical data related to EK and its associated factors and can serve as a reference for future studies, limitations must be considered. To illustrate, this study used a cross-sectional design, which precludes cause-and-effect relationships. Therefore, the generalizability of the findings should be interpreted with caution. Additionally, the study used a convenience sample technique because of constraints, such as time and money, which might have introduced some selection bias. However, the mentioned sampling approach was used to assess EK and its risk factors to lay the groundwork for protocol development and to inform practice but has not been applied in practice. Additionally, variation in interobserver reliability could have introduced some bias, although this was minimized using a coding manual. In addition, Cohen’s kappa was 0.80, indicating substantial agreement. Furthermore, the study used a validated tool corneal fluorescein staining for data collection and maintaining reliability through consistent use of equipment (e.g., a cobalt blue filter with an indirect ophthalmoscope). Finally, while some findings related to the regression model yielded wide confidence intervals, which could signal instability in the estimates, we believe our results are robust because we performed sample size calculation during planning to ensure that our sample was adequate to detect a meaningful effect. Some factors were not tested, such as the type of intubation (tracheostomy or endotracheal tube). Patients with endotracheal tube intubation have a higher incidence of ocular surface diseases than patients with tracheostomy [38]. Another limitation was that surviving patients were not followed after they were transferred to the floor.
In conclusion, the frequency of EK is high and underestimated in ventilated patients. Healthcare providers in ICUs should increase their awareness of the various risk factors involved in the development of EK, improving the quality of care of ventilated patients. Establishing an eye care protocol is very important and should be a part of daily care for sedated and ventilated patients in the ICU to reduce the incidence of EK.
KEY MESSAGES
▪ Exposure keratopathy (EK) is one of the most common ocular problems among ventilated patients in intensive care units (ICUs).
▪ Healthcare providers in ICUs prioritize life-threatening illnesses and stabilizing the patient’s medical status, with eye care a secondary issue.
▪ The frequency of EK is high and underestimated in ventilated patients.
▪ Healthcare providers in ICUs must raise their awareness of the various risk factors contributing to EK and improve the quality of care for ventilated patients.
Notes
CONFLICT OF INTEREST
No potential conflict of interest relevant to this article was reported.
FUNDING
None.
ACKNOWLEDGMENTS
None.
AUTHOR CONTRIBUTIONS
Conceptualization: ATS, SYNM. Methodology: ATS, SYNM, GAM. Formal analysis: ATS, SYNM. Data curation: ATS, AAF. Visualization: ATS, AAF, AAR. Project administration: GAM. Funding acquisition: SYNM. Writing – original draft: ATS. Writing – review & editing: SYNM, GAM, AAF, AAR. All authors read and agreed to the published version of the manuscript.