- Email: [email protected]
Incidence and risk factor evaluation of exposure keratopathy in critically ill patients: A cohort study
Journal of Critical Care 30 (2015) 400–404
Contents lists available at ScienceDirect
Journal of Critical Care journal homepage: www.jccjournal.org
Incidence and risk factor evaluation of exposure keratopathy in critically ill patients: A cohort study☆ Shilpa Kuruvilla, MS a, Jayanthi Peter, MS, DO, DNB a, Sarada David, MS, DO a, Prasanna Samuel Premkumar, M.Sc., PhD b,1, Kartik Ramakrishna, MBBS c,2, Lovely Thomas, MBBS DNB c, Manuel Vedakumar, MD c, John Victor Peter, MD, DNB, FRACP, FJFICM, FCICM c,⁎ a b c
Department of Ophthalmology, Schell hospital, Christian Medical College, Vellore, India Department of Biostatistics, Christian Medical College, Vellore, India Medical Intensive Care Unit, Christian Medical College, Vellore, India
a r t i c l e
i n f o
Keywords: Exposure keratopathy Critically ill Incidence Risk factors
a b s t r a c t Purpose: Recent emphasis on eye care in intensive care unit (ICU) patients has translated to eye assessment being part of routine care. In this setting, we determined the incidence, risk factors, and resolution time of exposure keratopathy. Methods: In this prospective cohort study, 301 patients were examined within 24 hours of ICU admission and subsequently daily by an ophthalmologist till death or discharge. Eyelid position, conjunctival and corneal changes, treatment, and outcome data were collected. Results: Admission diagnoses included febrile illnesses (35.2%) and respiratory failure (32.6%); 84.1% were ventilated. Forty-nine patients had exposure keratopathy (bilateral = 35, unilateral = 14) at admission; 35 patients developed new onset keratopathy (incidence 13.2%) 4.6 ± 2.6 days after ICU admission. In 67 patients, keratopathy was mild (punctate epithelial erosions). Macroepithelial defects (n = 9), stromal whitening with epithelial defect (n = 3), and stromal scar (n = 3) were infrequent. None developed microbial keratitis. On multivariate logistic regression analysis, eyelid position (odds ratio, 2.93; 95% confidence interval, 1.37-6.25), and ventilation duration (odds ratio, 1.11; 95% confidence interval, 1.04-1.19) were strongly associated with the development of keratopathy after ICU admission. Keratopathy resolved in 3.6 ± 4.5 days. Conclusions: Severe exposure keratopathy is infrequent in a protocolized ICU setting. Eyelid position and duration of ventilation are associated with exposure keratopathy. © 2014 Elsevier Inc. All rights reserved.
1. Introduction Exposure keratopathy refers to the drying of the cornea with subsequent epithelial breakdown [1]. This often occurs due to a failure of the eyelids to cover the globe, resulting in improper wetting of the ocular surface by tears [2]. Critically ill patients are particularly vulnerable to develop exposure keratopathy. Decreased tear formation [3], reduced blink rate [3,4], incomplete eye closure [5–7], decreased corneal reflex [7,8], and altered vascular permeability [7,8] predispose to exposure keratopathy. In the 1990s, exposure keratopathy was reported to occur in 20% to 40% of intensive care unit (ICU) patients [5,8,9]. The high ☆ There is no conflict of interest or financial disclosure for all the authors. No funding has been received for this study. ⁎ Corresponding author at: Medical Intensive Care Unit, Christian Medical College, Vellore 632004, India. Tel.: + 91 416 228 2693. E-mail address: [email protected] (J.V. Peter). 1 Current Affiliation: National Center for Immunization and Respiratory Diseases, CDC, Atlanta, GA, USA. 2 Current Affiliation: Department of Medicine, Unity Health System, Rochester, NY 14626, USA. http://dx.doi.org/10.1016/j.jcrc.2014.10.009 0883-9441/© 2014 Elsevier Inc. All rights reserved.
incidence prompted the development of protocols that incorporated eye care as part of assessment of organ function. This included the systematic daily evaluation of the eye to pick up early keratopathy, the liberal application of eye lubricants, and the use of moisture chambers [10–12] in high-risk patients. It is unclear if the above measures, which are now an integral part of management, have significantly impacted exposure keratopathy, although it is the perception among intensivists that this is so. On the other hand, advances in critical care have enabled the prolonged support of extremely sick patients, which in turn may influence the occurrence of exposure keratopathy. This study was thus aimed at determining the incidence of exposure keratopathy in patients admitted to the ICU. We also sought to study the risk factors associated with its development and the time to resolution of keratopathy.
2. Materials and methods This was a prospective, cohort study of consecutive patients admitted to the medical ICU of a 2300-bed tertiary care hospital for 3½ months.
S. Kuruvilla et al. / Journal of Critical Care 30 (2015) 400–404
The study was approved by the institution review board and ethics committee, and informed consent was obtained from patient or next of kin. Patients were recruited within 24 hours of ICU admission; those unwilling to participate or with ICU stay less than 24 hours were excluded. The eye care protocol in our unit includes daily eye examination by the ICU team (registrar and/or consultant) during each shift as part of assessment of organ dysfunction, use of eye lubricant ointments (fourth hourly) in patients with improper eye closure, and the use of goggles with or without taping in patients on neuromuscular blocking agents or improper eye closure. If exposure keratopathy is suspected by the ICU team at any stage, an ophthalmologist is called for assessment and advice on treatment. However, during the study period, in addition to the above protocol, patients were also assessed daily by the same ophthalmologist using the cobalt blue light of a portable slit-lamp bio-microscope (HEINE HSL 100, Hand-held slit lamp; Heine Optotechnik, Herrsching, Germany) after staining the ocular surface with fluorescein dye (disposable fluorescein strip moistened with a drop of hydroxypropyl methylcellulose). Eyelid position, conjunctival chemosis, and corneal changes were graded (Appendix A) as in previous studies [5,13]. The eye care provided for each patient was noted. Patient demographics, diagnosis, need and duration of ventilation, use and duration of sedation and neuromuscular blocking agents, and length of ICU stay and hospital stay were collected in preconstructed data abstraction forms. The standard analgo-sedation used in our unit was either morphine or fentanyl in combination with midazolam titrated to a Richmond Agitation Sedation Scale of − 2 to zero. Pancuronium was used as the sole neuromuscular blocking agent where indicated. Patients were followed up till death or discharge from ICU. Those who developed features of exposure keratopathy were treated with hourly Lacrigel (Sunways, Mumbai, India) lubricant eye ointment and taping or goggles as appropriate. Topical antibiotics were administered (ciprofloxacin) if there was clinical suspicion of microbial keratitis pending culture reports. Patients with exposure keratopathy were followed up till resolution of symptoms and signs.
401
2.1. Statistical methods The incidence of exposure keratopathy was calculated as the proportion developing keratopathy after ICU admission. Summary data were presented as mean (SD) or median (interquartile range [IQR]) as appropriate. The association between exposure keratopathy and risk factors was assessed using χ 2 or Fisher exact test for categorical and Wilcoxon rank sum test for continuous data. Furthermore, bivariate analysis was conducted using simple logistic regression to determine the association between risk factors and exposure keratopathy and expressed as odds ratio (OR) with 95% confidence interval (CI). Multivariate logistic regression analysis was then performed to identify independent risk factors for exposure keratopathy. Risk factors found to be significantly associated with keratopathy in the bivariate analyses was included in the multivariable analyses. All statistical analyses were performed using STATA 11 software (StataCorp, College Station, Tex). 3. Results Of the 314 patients admitted to the ICU during the study period, 13 were excluded (2 unwilling to participate; 11 died or were discharged within 24 hours of ICU admission). The study cohort comprised 301 patients with a mean (SD) age of 43 (16.9) years and Acute Physiology and Chronic Health Evaluation II (APACHE II) score of 18.2 (8.0) (Table 1). Main diagnoses included febrile illnesses (35.2%) and respiratory failure (32.6%); 84.1% needed ventilation, invasive (n = 213), noninvasive (n = 70), or both (n = 30). Sedation was used in 208 patients (69.1%) and neuromuscular blocking agents in 29 (9.6%) patients. The median duration of ICU stay was 4 (IQR, 3-7) days, and that of hospital stay was 10 (IQR, 6-16) days. Hospital mortality was 30.2%. The eyelid position was examined, graded (Appendix A), and recorded daily by the ophthalmologist. On first examination, 86.3%
Table 1 Demographics, treatment, and outcome of the entire cohort Parameter
All patients (n = 301)
No exposure keratopathy (n = 219)
EK diagnosed at or after ICU admissiona (n = 82)
EK after ICU admissionb (n = 35)
Age (y) Sex: male/female APACHE II scorec Reason for ICU admissiond Respiratory failure Hemodynamics Neurologic Poisoning Monitoring Number ventilated Type of ventilation Invasive ventilation Noninvasive ventilation Both Duration of ventilation (d)e Use of sedation (n) Duration of sedation (d)e Use of neuromuscular blocking drug (n) Need for dialysis (n) ICU length of stay (d)f Hospital length of stay (d)f ICU mortality (n) Hospital mortality (n)
43.4 (16.9) 170:131 18.2 (8.0)
45.1 (17.1) 121:98 17.1 (7.4)
38.7 (1.7) 49:33 20.9 (8.9)
41.3 (15.7) 22: 13 20.2 (8.3)
181 24 14 12 66 253 (84.1%)
129 16 12 5 54 174 (79.5)
52 8 2 7 12 79 (96.3%)
21 6 1 2 5 35 (100%)
213 (70.8%) 70 (23.3%) 30 (10%) 6.1 (5.3) 208 (69.1%) 4.9 (3.9) 29 (9.6%) 54 (17.9%) 4 (3-7) 10 (6-16) 81 (26.9%) 91 (30.2%)
134 (61.2%) 55 (25.1%) 15 (6.9%) 5.7 (4.6) 132 (60.3%) 4.7 (3.4) 7 (3.3%) 34 (15.5%) 4 (3-7) 10 (6-17) 33 (15.1%) 40 (18.3%)
79 (96.3%) 15 (18.3%) 15 (18.3%) 7.0 (6.6) 76 (92.7%) 5.1 (3.4) 22 (26.8%) 20 (24.2%) 5 (3-10) 8.5 (4.5-15) 48 (58.5%) 51 (62.2%)
35 (100%) 11 (31.4%) 11 (31.4%) 8.9 (8.3) 35 (100%) 5.6 (3.7) 11 (31.4%) 9 (25.7%) 7 (5-11) 10 (7-17) 25 (71.4%) 26 (74.3%)
EK indicates keratopathy; (n), number of patients. Values in parentheses indicate SD unless specified. a Eighty-two patients had exposure EK at any time during ICU admission. b Thirty-five patients developed exposure keratopathy after ICU admission. c Data available on 167 patients. d Data available on 297 patients. f Median (IQR). e The duration of ventilation and sedation were calculated only in those who were ventilated.
402
S. Kuruvilla et al. / Journal of Critical Care 30 (2015) 400–404
Table 2 Eyelid position on day 1 as well as the worst grade in the first 7 days after intensive care admission as recorded by the ophthalmologist Eye lid position grade
0 1 2 3 4 Total
Day 1
Table 3 Bivariate logistic regression analysis of factors associated with the development of exposure keratopathy after intensive care admission
Worst grade in the first 7 d
Variable
Crude ORa
P
95% CI
Age Sex Diagnostic category APACHE II score Eyelid position Conjunctival chemosis Duration of ventilationc Duration of sedation Use of neuromuscular blocking drug Need for dialysis ICU length of stay Hospital length of stay Hospital mortality
0.99 0.77 0.94 1.05 2.36 0.88 1.1 1.08 2.41 1.77 1.12 1.02 11.6
.26 .46 .52 .10 .013 .76 .004 .18 .06 .18 .002 .19 b.001
0.97-1.01 0.38-1.55 0.78-1.13 0.99-1.11 1.20-4.62b 0.37-2.06 1.03-1.17b 0.97-1.20 0.96-6.02 0.77-4.08 1.04-1.20b 0.99-1.04 5.10-26.5b
Right eye
Left eye
Right eye
Left eye
259 (86.3) 15 (5.0) 25 (8.3) 1 (0.3) 0 300
256 (85.3) 17 (5.7) 25 (8.3) 2 (0.7) 0 300
239 (79.7) 22 (7.3) 36 (12.0) 2 (0.7) 1 (0.3) 300
227 (75.7) 28 (9.3) 40 (13.3) 3 (1.0) 2 (0.7) 300
Data available only on 300 patients. Grade 0 indicates complete eyelid closure; Grade 1, only conjunctival exposure; grade 2, lower one fourth of the cornea exposed; grade 3, lower one half of the cornea exposed; grade 4, three fourth of the cornea exposed; grade 5, cornea fully exposed. All values in parentheses indicate percentages.
and 85.3% had full eyelid closure on day 1 in the right and left eyes, respectively (Table 2). Varying degrees of incomplete eye closure were observed in the remaining patients. Incomplete eyelid closure increased to 20.3% and 24.3% in the right and left eyes, respectively, in the first week of ICU admission (Table 2). Within the first 24 hours of ICU admission (ie, first ocular examination), 49 patients (16.3%) had evidence of exposure keratopathy (bilateral = 35, unilateral = 14). Of the 266 patients without exposure keratopathy (n = 252) or unilateral exposure keratopathy (n = 14) at admission, a further 35 patients (13.2%) developed “new-onset” keratopathy at 4.6 ± 2.6 days. The 35 patients included 2 patients with unilateral keratopathy at ICU admission who subsequently developed exposure keratopathy in the other eye. The incidence of exposure keratopathy after ICU admission was thus 13.2% in our cohort. In total, 82 patients had exposure keratopathy at any point during ICU stay. When graded (Appendix A), of the 82 patients, 67 (81.7%) had mild keratopathy (punctate epithelial erosions). Macroepithelial defects (n = 9), stromal whitening with epithelial defect (n = 3), and stromal scar (n = 3) were infrequent. Ocular treatment comprised lubricants (n = 112; 37.2%), taping of eyelids (n = 31; 10.3%), and goggles (n = 31; 10.3%). In addition, 5 patients were administered topical antibiotics on suspicion of microbial keratitis; cultures, however, were negative. Although keratopathy resolved in all patients by 3.6 ± 4.5 days, in 3 patients there was worsening of keratopathy in the initial 24 to 48 hours after diagnosis (by 1 grade in 2 patients and by 2 grades in 1 patient) prior to resolution. None developed sightthreatening complications or corneal perforation. In the 3 patients who developed corneal scarring, the scars were superficial and limited to the inferior one third of the cornea without interference with vision. Bivariate analysis was done using simple logistic regression on risk factors identified pre hoc for the development of keratopathy (Table 3). This was done only on the 266 patients who had either no exposure keratopathy at ICU admission (n = 252) or only unilateral exposure keratopathy (n = 14). In patients developing exposure keratopathy after ICU admission (n = 35), eyelid position was correlated with exposure keratopathy (OR, 2.36; 95% CI, 1.20-4.62; P = .013), whereas conjunctival chemosis was not (P = .76). The need for and duration of ventilation (invasive and noninvasive) and length of ICU stay were significantly (OR, 1.1; 95% CI, 1.03-1.17; P b .004) associated with the development of exposure keratopathy. The duration of sedation was not associated with exposure keratopathy (P = .18), although use of neuromuscular blocking agents tended (OR, 2.41; 95% CI, 0.96-6.02; P = .06) to be associated with the development of exposure keratopathy. Multivariate logistic regression analysis (Table 4) was done incorporating eyelid position, duration of ventilation, use of muscle relaxants, and duration of hospital stay. Eyelid position (OR, 2.69; 95% CI, 1.24-5.83; P = .012) and duration of ventilation (OR, 1.12; 95% CI, 1.03-1.22; P = .007) were strongly associated with the development of exposure keratopathy after ICU admission, whereas use of neuromuscular blocking agents was not (P = .107).
a Because the factors associated with the development of exposure keratopathy cannot be assessed on those who had bilateral exposure keratopathy at intensive care admission, this analysis was done only on the 266 patients who had no exposure keratopathy (n = 252) or unilateral exposure keratopathy (n = 14) at the time of intensive care admission. b Significant associations. c Need for ventilation predicts the development of exposure keratopathy perfectly.
Among the 82 patients noted to have keratopathy at any point during ICU stay, 48 died and 34 were discharged alive from ICU. The association of presence of exposure keratopathy and mortality was statistically significant (P b .0001). 4. Discussion In this prospective cohort study of 301 ICU patients, although 16.3% of patients had some evidence of exposure keratopathy at ICU admission, new-onset exposure keratopathy after ICU admission occurred only in 13.2%. The prevalence of exposure keratopathy was 27.3% (82/301). Among all patients with exposure keratopathy (n = 82), more than 80% had only punctate epithelial erosions; stromal whitening with epithelial defect and stromal scar were infrequent. None of the patients developed microbial keratitis. Keratopathy resolved fairly rapidly, in about 4 days. On multivariate logistic regression analysis, only eyelid position and ventilation duration were strongly associated with the development of keratopathy after ICU admission. Several studies have examined the incidence of exposure keratopathy in ICU patients (Table 5). Studies in the 1990s reported an incidence of 20% to 40% [5,8,9]. Subsequent to this, the incidence of exposure keratopathy has ranged from as high as 25% to 55.4% [14–16] to 8.7% to 13.4% [17–19]. The studies reporting a low incidence are limited either by the retrospective nature of the study or infrequent ocular examinations. It is possible that the use of an eye care algorithm [19] or an evidence-based eye care education [2] may reduce the incidence of exposure keratopathy. In the study by Suresh et al [19] that evaluated 34 patients, the prevalence of ocular surface abnormalities was 8.7% Table 4 Multivariate logistic regression analysis of factors associated with the development of exposure keratopathy Variable
Adjusted OR
P
95% CI
Eyelid position Ventilation duration Use of neuromuscular blocking drugs Hospital length of stay
2.69 1.12 1.74 0.99
.012 .007 .107 .585
1.24-5.83 1.03-1.22 0.89-3.42 0.95-1.03
If the duration of ventilation and length of hospital stay were considered as co-related variables, removal of the hospital length of stay from the analysis resulted in the eyelid position and ventilation duration continuing to be significantly associated with new onset exposure keratopathy, whereas use of neuromuscular blocking drugs was not associated. Replacement of hospital length of stay with duration of ventilation resulted in similar results where only the eyelid position and ventilation duration were associated with new onset exposure keratopathy.
S. Kuruvilla et al. / Journal of Critical Care 30 (2015) 400–404
403
Table 5 Summary of studies on exposure keratopathy in critically ill patients Study
No.
Study type
Assessment
Event rate
Interpretations
Current study (2014)
301
Prospective
Dailya
13.2%
Masoudi Alavi et al [18] Saritas et al [17] Jammal et al [15] Germano et al [16] McHugh et al [6]
87 40 74 53 18
Prospective Retrospective (stay N7 d) Prospective Prospective Prospective
Days 1 and 5b NA Dailyc Daily Twice weekly
13.8% 10% 57% 25% 37.5%
Desalu et al [14]
56
Prospective
Daily
55.4%
Suresh et al [19]
34
Prospective
Twice weekly
8.7%
Mercieca et al [5]
26
Prospective
At least weekly
42%
Hernandez and Mannis [9] Imanaka et al [8]
50 143
Random cohort Retrospective (stay ≥7 d)
NA NA
40% 20%
Risk factors eyelid position (P = .013), ventilation duration (P = .004), ICU stay (P = .002) 32.2% dry eyes and 13.8% corneal abrasion No association with mechanical ventilation, sedation, and use of inotropes The frequency of exposure keratopathy increased with lagophthalmos score Study in mechanically ventilated children; opened eyes and sepsis prognostic factors Sensitivity, specificity by ICU doctor 77.8%, 96.7%; associated with incomplete eye closure (P = .027) Ventilation duration, sedation and severity of illness but not eyelid position influences keratopathy Prevalence of ocular surface 8.7% when the algorithm was properly followed (23 patients) Keratopathy correlates with lagophthalmos, which in turn related to depth of sedation or paralysis Glasgow Coma Scale score b7 (P b .03), duration of stay beyond 7 d (P = .03) 35% with sedation, 39% neuromuscular blockade, inability to close eye higher incidence (P b .01)
NA indicates details not available. a Patients were assessed both by the ICU doctor and the ophthalmologist. b Assessment by ophthalmologist on the day of admission and on day 5. c First assessment by ophthalmologist was 1 to 5 days after commencing ventilation and subsequently daily.
when the algorithm was followed. However, in this study [19], assessments were done on patients only twice weekly, and hence, mild exposure keratopathy may have been missed. In a more recent study, exposure keratopathy was observed in 6 patients of 762 in the pretraining period and in 3 patients of 6196 in the posttraining period [2]. The relatively low incidence of new-onset exposure keratopathy of 13.2% in our study is consistent with the above observations that protocolized eye care may reduce the incidence not only of exposure keratopathy but also of severe exposure keratopathy or sightthreatening complications, which were infrequent in our study. Daily assessment by the ICU team and appropriate eye care is likely to have contributed to this. Furthermore, daily assessments by the ophthalmologists may have also resulted in early detection and treatment of exposure keratopathy in our patients. McHugh et al [6] showed that the sensitivity and specificity of detecting exposure keratopathy in ICU doctors were 77.8% and 96.7%, respectively, when compared with an ophthalmologist. Our finding of eyelid position being significantly associated with the development of exposure keratopathy is consistent with the earliest observations by Mercieca et al [5], who demonstrated that eyelid position was the most important factor determining the occurrence of exposure keratopathy. Other studies [8,15] have also demonstrated a relationship between lagophthalmos and exposure keratopathy. The need for mechanical ventilation and the duration of ventilation being associated with exposure keratopathy in our study are consistent with the observations by Hernandez and Mannis [9]. Although the same study [9] also found an association between sedation and neuromuscular paralysis and exposure keratopathy, we observed only a trend for an association between the use of neuromuscular blocking agents and exposure keratopathy on bivariate analysis but not sedation and exposure keratopathy. We also observed that patients who stayed longer in ICU developed exposure keratopathy, consistent with earlier observations [9]. The observation of an association between exposure keratopathy and mortality in our study is not surprising and probably reflects the fact that the sicker patients are more likely to develop exposure keratopathy and more likely to die, rather than a direct impact of exposure keratopathy on mortality. The major strength of this study is the large sample size. Previous studies on exposure keratopathy were limited by small numbers ranging from 26 to 143 patients (Table 5). Furthermore, all patients were examined daily by an ophthalmologist with a portable slit-lamp, this not being the case in most studies. However, such detailed evaluation limits external validity. In 4 studies [6,15,17,18], ophthalmologists
assessed the patients, albeit not daily. Because of practical difficulties, in our study, the first ocular examination by the ophthalmologist could not take place immediately after ICU admission. This may have overestimated the presence of exposure keratopathy at presentation to ICU. It should also be noted that daily examination by the ophthalmologist and a Hawthorne effect of the awareness of the conduct of the study among health care workers may have influenced the incidence and the course of the disease. However, this is likely to have impacted detection and not the course; as per protocol, all patients with exposure keratopathy are followed up daily by the ophthalmologist till resolution. Data on use of atypical antipsychotic agents, propofol, and dexmedetomidine as well fluid balance and use of vasoactive agents were not collected and correlated with the incidence of keratopathy, and this is a limitation. 5. Conclusions A protocolized management of systematic and frequent examination of the eye and the use of strategies to protect the eye (lubricants, goggles) was associated with a low incidence of exposure keratopathy of 13.2% without any occurrence of severe keratopathy. Its occurrence in patients with improper eyelid closure who need prolonged mechanical ventilation highlights the importance of meticulous care of the eye in such high-risk patients. Appendix A. Supplementary data Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.jcrc.2014.10.009. References [1] Friedman NJ, Kaiser PK. Cornea. In: Friedman NJ, Kaiser PK, editors. Essentials of ophthalmology. Elsevier Health Sciences; 2007. p. 180. [2] Demirel S, Cumurcu T, Firat P, Aydogan MS, Doganay S. Effective management of exposure keratopathy developed in intensive care units: the impact of an evidence based eye care education programme. Intensive Crit Care Nurs 2014;30:38–44. [3] Ousler 3rd GW, Hagberg KW, Schindelar M, Welch D, Abelson MB. The ocular protection index. Cornea 2008;27:509–13. [4] Patel V, Daya SM, Lake D, Malhotra R. Blink lagophthalmos and dry eye keratopathy in patients with non-facial palsy: clinical features and management with upper eyelid loading. Ophthalmology 2011;118:197–202. [5] Mercieca F, Suresh P, Morton A, Tullo A. Ocular surface disease in intensive care unit patients. Eye (Lond) 1999;13:231–6. [6] McHugh J, Alexander P, Kalhoro A, Ionides A. Screening for ocular surface disease in the intensive care unit. Eye (Lond) 2008;22:1465–8. [7] Rosenberg JB, Eisen LA. Eye care in the intensive care unit: narrative review and meta-analysis. Crit Care Med 2008;36:3151–5.
404
S. Kuruvilla et al. / Journal of Critical Care 30 (2015) 400–404
[8] Imanaka H, Taenaka N, Nakamura J, Aoyama K, Hosotani H. Ocular surface disorders in the critically ill. Anesth Analg 1997;85:343–6. [9] Hernandez EV, Mannis MJ. Superficial keratopathy in intensive care unit patients. Am J Ophthalmol 1997;12:212–6. [10] Cortese D, Capp L, McKinley S. Moisture chamber versus lubrication for the prevention of corneal epithelial breakdown. Am J Crit Care 1995;4:425–8. [11] Sivasankar S, Jasper S, Simon S, Jacob P, John G, Raju R. Eye care in ICU. Indian J Crit Care Med 2006;10:11–4. [12] Sorce LR, Hamilton SM, Gauvreau K, Mets MB, Hunter DG, Rahmani B, et al. Preventing corneal abrasions in critically ill children receiving neuromuscular blockade: a randomized controlled trial. Pediatr Crit Care Med 2009;10: 171–5. [13] Ezra DG, Lewis G, Healy M, Coombes A. Preventing exposure keratopathy in the critically ill: a prospective study comparing eye care regimes. Br J Ophthalmol 2005;89:1068–9.
[14] Desalu I, Akinsola F, Adekola O, Akinbami O, Kushimo O, Adefule-Ositelu A. Ocular surface disorder in intensive care unit patients in a sub-Saharan teaching hospital. Internet J Emerg Intensive Care Med 2007;11(1). [15] Jammal H, Khader Y, Shihadeh W, Abadneh L, Alijzawi G, AlQasem A. Exposure keratopathy in sedated and ventilated patients. J Crit Care 2012;27:537–41. [16] Germano EM, Mello MJ, Sena DF, Correia JB, Amorim MM. Incidence and risk factors of corneal epithelial defects in mechanically ventilated children. Crit Care Med 2009; 37:1097–100. [17] Saritas TB, Bozkurt B, Simsek B, Cakmak Z, Ozdemir M, Yosunkaya A. Ocular surface disorders in intensive care unit patients. Scientific World Journal 2013;182038. http://dx.doi.org/10.1155/2013/182038. [18] Masoudi Alavi N, Sharifitabar Z, Shaeri M, Abib-Hajbaghery M. An audit of eye dryness and corneal abrasion in ICU patients in Iran. Nurs Crit Care 2014;19:73–7. [19] Suresh P, Mercieca F, Morton A, Tullo AB. Eye care for the critically ill. Intensive Care Med 2000;26:162–6.
Contents lists available at ScienceDirect
Journal of Critical Care journal homepage: www.jccjournal.org
Incidence and risk factor evaluation of exposure keratopathy in critically ill patients: A cohort study☆ Shilpa Kuruvilla, MS a, Jayanthi Peter, MS, DO, DNB a, Sarada David, MS, DO a, Prasanna Samuel Premkumar, M.Sc., PhD b,1, Kartik Ramakrishna, MBBS c,2, Lovely Thomas, MBBS DNB c, Manuel Vedakumar, MD c, John Victor Peter, MD, DNB, FRACP, FJFICM, FCICM c,⁎ a b c
Department of Ophthalmology, Schell hospital, Christian Medical College, Vellore, India Department of Biostatistics, Christian Medical College, Vellore, India Medical Intensive Care Unit, Christian Medical College, Vellore, India
a r t i c l e
i n f o
Keywords: Exposure keratopathy Critically ill Incidence Risk factors
a b s t r a c t Purpose: Recent emphasis on eye care in intensive care unit (ICU) patients has translated to eye assessment being part of routine care. In this setting, we determined the incidence, risk factors, and resolution time of exposure keratopathy. Methods: In this prospective cohort study, 301 patients were examined within 24 hours of ICU admission and subsequently daily by an ophthalmologist till death or discharge. Eyelid position, conjunctival and corneal changes, treatment, and outcome data were collected. Results: Admission diagnoses included febrile illnesses (35.2%) and respiratory failure (32.6%); 84.1% were ventilated. Forty-nine patients had exposure keratopathy (bilateral = 35, unilateral = 14) at admission; 35 patients developed new onset keratopathy (incidence 13.2%) 4.6 ± 2.6 days after ICU admission. In 67 patients, keratopathy was mild (punctate epithelial erosions). Macroepithelial defects (n = 9), stromal whitening with epithelial defect (n = 3), and stromal scar (n = 3) were infrequent. None developed microbial keratitis. On multivariate logistic regression analysis, eyelid position (odds ratio, 2.93; 95% confidence interval, 1.37-6.25), and ventilation duration (odds ratio, 1.11; 95% confidence interval, 1.04-1.19) were strongly associated with the development of keratopathy after ICU admission. Keratopathy resolved in 3.6 ± 4.5 days. Conclusions: Severe exposure keratopathy is infrequent in a protocolized ICU setting. Eyelid position and duration of ventilation are associated with exposure keratopathy. © 2014 Elsevier Inc. All rights reserved.
1. Introduction Exposure keratopathy refers to the drying of the cornea with subsequent epithelial breakdown [1]. This often occurs due to a failure of the eyelids to cover the globe, resulting in improper wetting of the ocular surface by tears [2]. Critically ill patients are particularly vulnerable to develop exposure keratopathy. Decreased tear formation [3], reduced blink rate [3,4], incomplete eye closure [5–7], decreased corneal reflex [7,8], and altered vascular permeability [7,8] predispose to exposure keratopathy. In the 1990s, exposure keratopathy was reported to occur in 20% to 40% of intensive care unit (ICU) patients [5,8,9]. The high ☆ There is no conflict of interest or financial disclosure for all the authors. No funding has been received for this study. ⁎ Corresponding author at: Medical Intensive Care Unit, Christian Medical College, Vellore 632004, India. Tel.: + 91 416 228 2693. E-mail address: [email protected] (J.V. Peter). 1 Current Affiliation: National Center for Immunization and Respiratory Diseases, CDC, Atlanta, GA, USA. 2 Current Affiliation: Department of Medicine, Unity Health System, Rochester, NY 14626, USA. http://dx.doi.org/10.1016/j.jcrc.2014.10.009 0883-9441/© 2014 Elsevier Inc. All rights reserved.
incidence prompted the development of protocols that incorporated eye care as part of assessment of organ function. This included the systematic daily evaluation of the eye to pick up early keratopathy, the liberal application of eye lubricants, and the use of moisture chambers [10–12] in high-risk patients. It is unclear if the above measures, which are now an integral part of management, have significantly impacted exposure keratopathy, although it is the perception among intensivists that this is so. On the other hand, advances in critical care have enabled the prolonged support of extremely sick patients, which in turn may influence the occurrence of exposure keratopathy. This study was thus aimed at determining the incidence of exposure keratopathy in patients admitted to the ICU. We also sought to study the risk factors associated with its development and the time to resolution of keratopathy.
2. Materials and methods This was a prospective, cohort study of consecutive patients admitted to the medical ICU of a 2300-bed tertiary care hospital for 3½ months.
S. Kuruvilla et al. / Journal of Critical Care 30 (2015) 400–404
The study was approved by the institution review board and ethics committee, and informed consent was obtained from patient or next of kin. Patients were recruited within 24 hours of ICU admission; those unwilling to participate or with ICU stay less than 24 hours were excluded. The eye care protocol in our unit includes daily eye examination by the ICU team (registrar and/or consultant) during each shift as part of assessment of organ dysfunction, use of eye lubricant ointments (fourth hourly) in patients with improper eye closure, and the use of goggles with or without taping in patients on neuromuscular blocking agents or improper eye closure. If exposure keratopathy is suspected by the ICU team at any stage, an ophthalmologist is called for assessment and advice on treatment. However, during the study period, in addition to the above protocol, patients were also assessed daily by the same ophthalmologist using the cobalt blue light of a portable slit-lamp bio-microscope (HEINE HSL 100, Hand-held slit lamp; Heine Optotechnik, Herrsching, Germany) after staining the ocular surface with fluorescein dye (disposable fluorescein strip moistened with a drop of hydroxypropyl methylcellulose). Eyelid position, conjunctival chemosis, and corneal changes were graded (Appendix A) as in previous studies [5,13]. The eye care provided for each patient was noted. Patient demographics, diagnosis, need and duration of ventilation, use and duration of sedation and neuromuscular blocking agents, and length of ICU stay and hospital stay were collected in preconstructed data abstraction forms. The standard analgo-sedation used in our unit was either morphine or fentanyl in combination with midazolam titrated to a Richmond Agitation Sedation Scale of − 2 to zero. Pancuronium was used as the sole neuromuscular blocking agent where indicated. Patients were followed up till death or discharge from ICU. Those who developed features of exposure keratopathy were treated with hourly Lacrigel (Sunways, Mumbai, India) lubricant eye ointment and taping or goggles as appropriate. Topical antibiotics were administered (ciprofloxacin) if there was clinical suspicion of microbial keratitis pending culture reports. Patients with exposure keratopathy were followed up till resolution of symptoms and signs.
401
2.1. Statistical methods The incidence of exposure keratopathy was calculated as the proportion developing keratopathy after ICU admission. Summary data were presented as mean (SD) or median (interquartile range [IQR]) as appropriate. The association between exposure keratopathy and risk factors was assessed using χ 2 or Fisher exact test for categorical and Wilcoxon rank sum test for continuous data. Furthermore, bivariate analysis was conducted using simple logistic regression to determine the association between risk factors and exposure keratopathy and expressed as odds ratio (OR) with 95% confidence interval (CI). Multivariate logistic regression analysis was then performed to identify independent risk factors for exposure keratopathy. Risk factors found to be significantly associated with keratopathy in the bivariate analyses was included in the multivariable analyses. All statistical analyses were performed using STATA 11 software (StataCorp, College Station, Tex). 3. Results Of the 314 patients admitted to the ICU during the study period, 13 were excluded (2 unwilling to participate; 11 died or were discharged within 24 hours of ICU admission). The study cohort comprised 301 patients with a mean (SD) age of 43 (16.9) years and Acute Physiology and Chronic Health Evaluation II (APACHE II) score of 18.2 (8.0) (Table 1). Main diagnoses included febrile illnesses (35.2%) and respiratory failure (32.6%); 84.1% needed ventilation, invasive (n = 213), noninvasive (n = 70), or both (n = 30). Sedation was used in 208 patients (69.1%) and neuromuscular blocking agents in 29 (9.6%) patients. The median duration of ICU stay was 4 (IQR, 3-7) days, and that of hospital stay was 10 (IQR, 6-16) days. Hospital mortality was 30.2%. The eyelid position was examined, graded (Appendix A), and recorded daily by the ophthalmologist. On first examination, 86.3%
Table 1 Demographics, treatment, and outcome of the entire cohort Parameter
All patients (n = 301)
No exposure keratopathy (n = 219)
EK diagnosed at or after ICU admissiona (n = 82)
EK after ICU admissionb (n = 35)
Age (y) Sex: male/female APACHE II scorec Reason for ICU admissiond Respiratory failure Hemodynamics Neurologic Poisoning Monitoring Number ventilated Type of ventilation Invasive ventilation Noninvasive ventilation Both Duration of ventilation (d)e Use of sedation (n) Duration of sedation (d)e Use of neuromuscular blocking drug (n) Need for dialysis (n) ICU length of stay (d)f Hospital length of stay (d)f ICU mortality (n) Hospital mortality (n)
43.4 (16.9) 170:131 18.2 (8.0)
45.1 (17.1) 121:98 17.1 (7.4)
38.7 (1.7) 49:33 20.9 (8.9)
41.3 (15.7) 22: 13 20.2 (8.3)
181 24 14 12 66 253 (84.1%)
129 16 12 5 54 174 (79.5)
52 8 2 7 12 79 (96.3%)
21 6 1 2 5 35 (100%)
213 (70.8%) 70 (23.3%) 30 (10%) 6.1 (5.3) 208 (69.1%) 4.9 (3.9) 29 (9.6%) 54 (17.9%) 4 (3-7) 10 (6-16) 81 (26.9%) 91 (30.2%)
134 (61.2%) 55 (25.1%) 15 (6.9%) 5.7 (4.6) 132 (60.3%) 4.7 (3.4) 7 (3.3%) 34 (15.5%) 4 (3-7) 10 (6-17) 33 (15.1%) 40 (18.3%)
79 (96.3%) 15 (18.3%) 15 (18.3%) 7.0 (6.6) 76 (92.7%) 5.1 (3.4) 22 (26.8%) 20 (24.2%) 5 (3-10) 8.5 (4.5-15) 48 (58.5%) 51 (62.2%)
35 (100%) 11 (31.4%) 11 (31.4%) 8.9 (8.3) 35 (100%) 5.6 (3.7) 11 (31.4%) 9 (25.7%) 7 (5-11) 10 (7-17) 25 (71.4%) 26 (74.3%)
EK indicates keratopathy; (n), number of patients. Values in parentheses indicate SD unless specified. a Eighty-two patients had exposure EK at any time during ICU admission. b Thirty-five patients developed exposure keratopathy after ICU admission. c Data available on 167 patients. d Data available on 297 patients. f Median (IQR). e The duration of ventilation and sedation were calculated only in those who were ventilated.
402
S. Kuruvilla et al. / Journal of Critical Care 30 (2015) 400–404
Table 2 Eyelid position on day 1 as well as the worst grade in the first 7 days after intensive care admission as recorded by the ophthalmologist Eye lid position grade
0 1 2 3 4 Total
Day 1
Table 3 Bivariate logistic regression analysis of factors associated with the development of exposure keratopathy after intensive care admission
Worst grade in the first 7 d
Variable
Crude ORa
P
95% CI
Age Sex Diagnostic category APACHE II score Eyelid position Conjunctival chemosis Duration of ventilationc Duration of sedation Use of neuromuscular blocking drug Need for dialysis ICU length of stay Hospital length of stay Hospital mortality
0.99 0.77 0.94 1.05 2.36 0.88 1.1 1.08 2.41 1.77 1.12 1.02 11.6
.26 .46 .52 .10 .013 .76 .004 .18 .06 .18 .002 .19 b.001
0.97-1.01 0.38-1.55 0.78-1.13 0.99-1.11 1.20-4.62b 0.37-2.06 1.03-1.17b 0.97-1.20 0.96-6.02 0.77-4.08 1.04-1.20b 0.99-1.04 5.10-26.5b
Right eye
Left eye
Right eye
Left eye
259 (86.3) 15 (5.0) 25 (8.3) 1 (0.3) 0 300
256 (85.3) 17 (5.7) 25 (8.3) 2 (0.7) 0 300
239 (79.7) 22 (7.3) 36 (12.0) 2 (0.7) 1 (0.3) 300
227 (75.7) 28 (9.3) 40 (13.3) 3 (1.0) 2 (0.7) 300
Data available only on 300 patients. Grade 0 indicates complete eyelid closure; Grade 1, only conjunctival exposure; grade 2, lower one fourth of the cornea exposed; grade 3, lower one half of the cornea exposed; grade 4, three fourth of the cornea exposed; grade 5, cornea fully exposed. All values in parentheses indicate percentages.
and 85.3% had full eyelid closure on day 1 in the right and left eyes, respectively (Table 2). Varying degrees of incomplete eye closure were observed in the remaining patients. Incomplete eyelid closure increased to 20.3% and 24.3% in the right and left eyes, respectively, in the first week of ICU admission (Table 2). Within the first 24 hours of ICU admission (ie, first ocular examination), 49 patients (16.3%) had evidence of exposure keratopathy (bilateral = 35, unilateral = 14). Of the 266 patients without exposure keratopathy (n = 252) or unilateral exposure keratopathy (n = 14) at admission, a further 35 patients (13.2%) developed “new-onset” keratopathy at 4.6 ± 2.6 days. The 35 patients included 2 patients with unilateral keratopathy at ICU admission who subsequently developed exposure keratopathy in the other eye. The incidence of exposure keratopathy after ICU admission was thus 13.2% in our cohort. In total, 82 patients had exposure keratopathy at any point during ICU stay. When graded (Appendix A), of the 82 patients, 67 (81.7%) had mild keratopathy (punctate epithelial erosions). Macroepithelial defects (n = 9), stromal whitening with epithelial defect (n = 3), and stromal scar (n = 3) were infrequent. Ocular treatment comprised lubricants (n = 112; 37.2%), taping of eyelids (n = 31; 10.3%), and goggles (n = 31; 10.3%). In addition, 5 patients were administered topical antibiotics on suspicion of microbial keratitis; cultures, however, were negative. Although keratopathy resolved in all patients by 3.6 ± 4.5 days, in 3 patients there was worsening of keratopathy in the initial 24 to 48 hours after diagnosis (by 1 grade in 2 patients and by 2 grades in 1 patient) prior to resolution. None developed sightthreatening complications or corneal perforation. In the 3 patients who developed corneal scarring, the scars were superficial and limited to the inferior one third of the cornea without interference with vision. Bivariate analysis was done using simple logistic regression on risk factors identified pre hoc for the development of keratopathy (Table 3). This was done only on the 266 patients who had either no exposure keratopathy at ICU admission (n = 252) or only unilateral exposure keratopathy (n = 14). In patients developing exposure keratopathy after ICU admission (n = 35), eyelid position was correlated with exposure keratopathy (OR, 2.36; 95% CI, 1.20-4.62; P = .013), whereas conjunctival chemosis was not (P = .76). The need for and duration of ventilation (invasive and noninvasive) and length of ICU stay were significantly (OR, 1.1; 95% CI, 1.03-1.17; P b .004) associated with the development of exposure keratopathy. The duration of sedation was not associated with exposure keratopathy (P = .18), although use of neuromuscular blocking agents tended (OR, 2.41; 95% CI, 0.96-6.02; P = .06) to be associated with the development of exposure keratopathy. Multivariate logistic regression analysis (Table 4) was done incorporating eyelid position, duration of ventilation, use of muscle relaxants, and duration of hospital stay. Eyelid position (OR, 2.69; 95% CI, 1.24-5.83; P = .012) and duration of ventilation (OR, 1.12; 95% CI, 1.03-1.22; P = .007) were strongly associated with the development of exposure keratopathy after ICU admission, whereas use of neuromuscular blocking agents was not (P = .107).
a Because the factors associated with the development of exposure keratopathy cannot be assessed on those who had bilateral exposure keratopathy at intensive care admission, this analysis was done only on the 266 patients who had no exposure keratopathy (n = 252) or unilateral exposure keratopathy (n = 14) at the time of intensive care admission. b Significant associations. c Need for ventilation predicts the development of exposure keratopathy perfectly.
Among the 82 patients noted to have keratopathy at any point during ICU stay, 48 died and 34 were discharged alive from ICU. The association of presence of exposure keratopathy and mortality was statistically significant (P b .0001). 4. Discussion In this prospective cohort study of 301 ICU patients, although 16.3% of patients had some evidence of exposure keratopathy at ICU admission, new-onset exposure keratopathy after ICU admission occurred only in 13.2%. The prevalence of exposure keratopathy was 27.3% (82/301). Among all patients with exposure keratopathy (n = 82), more than 80% had only punctate epithelial erosions; stromal whitening with epithelial defect and stromal scar were infrequent. None of the patients developed microbial keratitis. Keratopathy resolved fairly rapidly, in about 4 days. On multivariate logistic regression analysis, only eyelid position and ventilation duration were strongly associated with the development of keratopathy after ICU admission. Several studies have examined the incidence of exposure keratopathy in ICU patients (Table 5). Studies in the 1990s reported an incidence of 20% to 40% [5,8,9]. Subsequent to this, the incidence of exposure keratopathy has ranged from as high as 25% to 55.4% [14–16] to 8.7% to 13.4% [17–19]. The studies reporting a low incidence are limited either by the retrospective nature of the study or infrequent ocular examinations. It is possible that the use of an eye care algorithm [19] or an evidence-based eye care education [2] may reduce the incidence of exposure keratopathy. In the study by Suresh et al [19] that evaluated 34 patients, the prevalence of ocular surface abnormalities was 8.7% Table 4 Multivariate logistic regression analysis of factors associated with the development of exposure keratopathy Variable
Adjusted OR
P
95% CI
Eyelid position Ventilation duration Use of neuromuscular blocking drugs Hospital length of stay
2.69 1.12 1.74 0.99
.012 .007 .107 .585
1.24-5.83 1.03-1.22 0.89-3.42 0.95-1.03
If the duration of ventilation and length of hospital stay were considered as co-related variables, removal of the hospital length of stay from the analysis resulted in the eyelid position and ventilation duration continuing to be significantly associated with new onset exposure keratopathy, whereas use of neuromuscular blocking drugs was not associated. Replacement of hospital length of stay with duration of ventilation resulted in similar results where only the eyelid position and ventilation duration were associated with new onset exposure keratopathy.
S. Kuruvilla et al. / Journal of Critical Care 30 (2015) 400–404
403
Table 5 Summary of studies on exposure keratopathy in critically ill patients Study
No.
Study type
Assessment
Event rate
Interpretations
Current study (2014)
301
Prospective
Dailya
13.2%
Masoudi Alavi et al [18] Saritas et al [17] Jammal et al [15] Germano et al [16] McHugh et al [6]
87 40 74 53 18
Prospective Retrospective (stay N7 d) Prospective Prospective Prospective
Days 1 and 5b NA Dailyc Daily Twice weekly
13.8% 10% 57% 25% 37.5%
Desalu et al [14]
56
Prospective
Daily
55.4%
Suresh et al [19]
34
Prospective
Twice weekly
8.7%
Mercieca et al [5]
26
Prospective
At least weekly
42%
Hernandez and Mannis [9] Imanaka et al [8]
50 143
Random cohort Retrospective (stay ≥7 d)
NA NA
40% 20%
Risk factors eyelid position (P = .013), ventilation duration (P = .004), ICU stay (P = .002) 32.2% dry eyes and 13.8% corneal abrasion No association with mechanical ventilation, sedation, and use of inotropes The frequency of exposure keratopathy increased with lagophthalmos score Study in mechanically ventilated children; opened eyes and sepsis prognostic factors Sensitivity, specificity by ICU doctor 77.8%, 96.7%; associated with incomplete eye closure (P = .027) Ventilation duration, sedation and severity of illness but not eyelid position influences keratopathy Prevalence of ocular surface 8.7% when the algorithm was properly followed (23 patients) Keratopathy correlates with lagophthalmos, which in turn related to depth of sedation or paralysis Glasgow Coma Scale score b7 (P b .03), duration of stay beyond 7 d (P = .03) 35% with sedation, 39% neuromuscular blockade, inability to close eye higher incidence (P b .01)
NA indicates details not available. a Patients were assessed both by the ICU doctor and the ophthalmologist. b Assessment by ophthalmologist on the day of admission and on day 5. c First assessment by ophthalmologist was 1 to 5 days after commencing ventilation and subsequently daily.
when the algorithm was followed. However, in this study [19], assessments were done on patients only twice weekly, and hence, mild exposure keratopathy may have been missed. In a more recent study, exposure keratopathy was observed in 6 patients of 762 in the pretraining period and in 3 patients of 6196 in the posttraining period [2]. The relatively low incidence of new-onset exposure keratopathy of 13.2% in our study is consistent with the above observations that protocolized eye care may reduce the incidence not only of exposure keratopathy but also of severe exposure keratopathy or sightthreatening complications, which were infrequent in our study. Daily assessment by the ICU team and appropriate eye care is likely to have contributed to this. Furthermore, daily assessments by the ophthalmologists may have also resulted in early detection and treatment of exposure keratopathy in our patients. McHugh et al [6] showed that the sensitivity and specificity of detecting exposure keratopathy in ICU doctors were 77.8% and 96.7%, respectively, when compared with an ophthalmologist. Our finding of eyelid position being significantly associated with the development of exposure keratopathy is consistent with the earliest observations by Mercieca et al [5], who demonstrated that eyelid position was the most important factor determining the occurrence of exposure keratopathy. Other studies [8,15] have also demonstrated a relationship between lagophthalmos and exposure keratopathy. The need for mechanical ventilation and the duration of ventilation being associated with exposure keratopathy in our study are consistent with the observations by Hernandez and Mannis [9]. Although the same study [9] also found an association between sedation and neuromuscular paralysis and exposure keratopathy, we observed only a trend for an association between the use of neuromuscular blocking agents and exposure keratopathy on bivariate analysis but not sedation and exposure keratopathy. We also observed that patients who stayed longer in ICU developed exposure keratopathy, consistent with earlier observations [9]. The observation of an association between exposure keratopathy and mortality in our study is not surprising and probably reflects the fact that the sicker patients are more likely to develop exposure keratopathy and more likely to die, rather than a direct impact of exposure keratopathy on mortality. The major strength of this study is the large sample size. Previous studies on exposure keratopathy were limited by small numbers ranging from 26 to 143 patients (Table 5). Furthermore, all patients were examined daily by an ophthalmologist with a portable slit-lamp, this not being the case in most studies. However, such detailed evaluation limits external validity. In 4 studies [6,15,17,18], ophthalmologists
assessed the patients, albeit not daily. Because of practical difficulties, in our study, the first ocular examination by the ophthalmologist could not take place immediately after ICU admission. This may have overestimated the presence of exposure keratopathy at presentation to ICU. It should also be noted that daily examination by the ophthalmologist and a Hawthorne effect of the awareness of the conduct of the study among health care workers may have influenced the incidence and the course of the disease. However, this is likely to have impacted detection and not the course; as per protocol, all patients with exposure keratopathy are followed up daily by the ophthalmologist till resolution. Data on use of atypical antipsychotic agents, propofol, and dexmedetomidine as well fluid balance and use of vasoactive agents were not collected and correlated with the incidence of keratopathy, and this is a limitation. 5. Conclusions A protocolized management of systematic and frequent examination of the eye and the use of strategies to protect the eye (lubricants, goggles) was associated with a low incidence of exposure keratopathy of 13.2% without any occurrence of severe keratopathy. Its occurrence in patients with improper eyelid closure who need prolonged mechanical ventilation highlights the importance of meticulous care of the eye in such high-risk patients. Appendix A. Supplementary data Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.jcrc.2014.10.009. References [1] Friedman NJ, Kaiser PK. Cornea. In: Friedman NJ, Kaiser PK, editors. Essentials of ophthalmology. Elsevier Health Sciences; 2007. p. 180. [2] Demirel S, Cumurcu T, Firat P, Aydogan MS, Doganay S. Effective management of exposure keratopathy developed in intensive care units: the impact of an evidence based eye care education programme. Intensive Crit Care Nurs 2014;30:38–44. [3] Ousler 3rd GW, Hagberg KW, Schindelar M, Welch D, Abelson MB. The ocular protection index. Cornea 2008;27:509–13. [4] Patel V, Daya SM, Lake D, Malhotra R. Blink lagophthalmos and dry eye keratopathy in patients with non-facial palsy: clinical features and management with upper eyelid loading. Ophthalmology 2011;118:197–202. [5] Mercieca F, Suresh P, Morton A, Tullo A. Ocular surface disease in intensive care unit patients. Eye (Lond) 1999;13:231–6. [6] McHugh J, Alexander P, Kalhoro A, Ionides A. Screening for ocular surface disease in the intensive care unit. Eye (Lond) 2008;22:1465–8. [7] Rosenberg JB, Eisen LA. Eye care in the intensive care unit: narrative review and meta-analysis. Crit Care Med 2008;36:3151–5.
404
S. Kuruvilla et al. / Journal of Critical Care 30 (2015) 400–404
[8] Imanaka H, Taenaka N, Nakamura J, Aoyama K, Hosotani H. Ocular surface disorders in the critically ill. Anesth Analg 1997;85:343–6. [9] Hernandez EV, Mannis MJ. Superficial keratopathy in intensive care unit patients. Am J Ophthalmol 1997;12:212–6. [10] Cortese D, Capp L, McKinley S. Moisture chamber versus lubrication for the prevention of corneal epithelial breakdown. Am J Crit Care 1995;4:425–8. [11] Sivasankar S, Jasper S, Simon S, Jacob P, John G, Raju R. Eye care in ICU. Indian J Crit Care Med 2006;10:11–4. [12] Sorce LR, Hamilton SM, Gauvreau K, Mets MB, Hunter DG, Rahmani B, et al. Preventing corneal abrasions in critically ill children receiving neuromuscular blockade: a randomized controlled trial. Pediatr Crit Care Med 2009;10: 171–5. [13] Ezra DG, Lewis G, Healy M, Coombes A. Preventing exposure keratopathy in the critically ill: a prospective study comparing eye care regimes. Br J Ophthalmol 2005;89:1068–9.
[14] Desalu I, Akinsola F, Adekola O, Akinbami O, Kushimo O, Adefule-Ositelu A. Ocular surface disorder in intensive care unit patients in a sub-Saharan teaching hospital. Internet J Emerg Intensive Care Med 2007;11(1). [15] Jammal H, Khader Y, Shihadeh W, Abadneh L, Alijzawi G, AlQasem A. Exposure keratopathy in sedated and ventilated patients. J Crit Care 2012;27:537–41. [16] Germano EM, Mello MJ, Sena DF, Correia JB, Amorim MM. Incidence and risk factors of corneal epithelial defects in mechanically ventilated children. Crit Care Med 2009; 37:1097–100. [17] Saritas TB, Bozkurt B, Simsek B, Cakmak Z, Ozdemir M, Yosunkaya A. Ocular surface disorders in intensive care unit patients. Scientific World Journal 2013;182038. http://dx.doi.org/10.1155/2013/182038. [18] Masoudi Alavi N, Sharifitabar Z, Shaeri M, Abib-Hajbaghery M. An audit of eye dryness and corneal abrasion in ICU patients in Iran. Nurs Crit Care 2014;19:73–7. [19] Suresh P, Mercieca F, Morton A, Tullo AB. Eye care for the critically ill. Intensive Care Med 2000;26:162–6.