Introduction
Retinal hemorrhage (RH) is a common finding in child abuse/shaken baby syndrome, but the presence of RH may also be considered with cardiopulmonary resuscitation due to thoracic compression, trauma, intracranial hemorrhages, raised intracranial pressure, and a wide range of illnesses such as blood disorders, infections (i.e., cytomegalovirus, herpes simplex virus, rickettsia, toxoplasmosis, endocarditis), and seizures1-4. In addition, birth-related RH is common in neonates due to spontaneous vaginal delivery (25%) and delivery with forceps and vacuum (43%-57%), while the lowest rate of RH occurs with cesarean section (7%)5,6. RH is a condition that has important medical, social and legal consequences. The current practice of determining RH in critically ill pediatric population in the developing world is not well characterized, and existing studies describing RH mostly focus on abusive head trauma7,8, and the presence of RH raises the question of preceding trauma/child abuse9.
Some studies state that RH might result from resuscitative efforts due to chest compression10,11. According to Agrawal et al., RHs were seen in critically ill children with a prevalence of 15.1%, and mainly (66%) were mild. Severe multilayered RH is uncommon (3.7%) and is only seen in patients with severe head trauma, fatal coagulopathy, sepsis, or a combination of these factors12. The aim of this prospective observational study was to define the prevalence, distribution, and extent of RH in critically ill children, defining demographic characteristics, clinical features and outcome of patients with RH, excluding newborns and patients with trauma. Furthermore, we also aimed to describe the impact of age, gender, prevailing diagnosis at admission, risk factors of respiratory failure, requirement of mechanical ventilation, presence of an endotracheal/tracheostomy tube, arrhythmia, hyperkalemia, pneumothorax, sepsis, heart failure, hypo/hypertension, presence of seizures, and presence of coagulation disorders on this distribution as secondary results.
Materials and Methods
Upon approval from the local Ethics Committee, the study was performed in a pediatric intensive care unit (PICU) which has 14 pediatric intensive care beds and is one of the largest referral centers for multidisciplinary tertiary care. It admits nearly 500-600 patients per year from all pediatric specialties. Patients between 1.5 month and 18 years of age with no historical or physical evidence of preceding trauma were included in the study. Data were collected on all patients with various diagnoses who had been admitted to our unit consecutively over a time period of one year, from April 2013 to April 2014.
During this period, 498 children were admitted to our unit. Among them, 150 patients were identified to meet the study inclusion criteria and their parents provided a written informed consent for the study. Demographic data such as age, gender, admission diagnosis, risk factors such as respiratory failure, requirement of mechanical ventilation, presence of an endotracheal/tracheostomy tube, cardiopulmonary resuscitation (CPR), arrhythmia, hyperkalemia, pneumothorax, sepsis, heart failure, effects of systolic and diastolic blood pressure values (mm Hg), presence of seizures, presence of coagulation disorders and platelet count, prothrombin time (PT) and partial thromboplastin time (PTT) levels were collected for statistical analysis. Coagulopathy was defined as PT of more than 3 seconds above normal, and/or twice normal PTT, and/or platelet count <50 000/mm3 13. Blood potassium level was determined according to its reference range to define hypokalemia or hyperkalemia14. In addition, hypotension and hypertension were defined according to its reference ranges according to age and gender14.
We classified RH by using laterality (unilateral or bilateral), severity (mild when less than 5 RH were identified, moderate when there were 5-20 RH, and severe when there were more than 20 RH), and single or multilayered based on their location in the retina5. All data were gathered and documented prospectively for statistical evaluation. All study patients underwent history taking, physical examination and routine laboratory investigations according to diagnosis, and these pediatric intensive care admissions had dilated retinal fundoscopic examination by an experienced ophthalmologist using an indirect ophthalmoscope. Eye examination was performed as soon as possible within the first 24-48 hours of admission. All children were monitored for vital functions such as heart and respiratory rate, blood pressure, pulse-oximeter, electrocardiogram (ECG), and urine output. Medical records of the patients were evaluated and prospective chart reviews were performed. Patients younger than 1.5 months (to exclude birth-related RH) and those older than 18 years, patients with suspected/proven abusive head trauma or penetrating eye trauma, all other trauma patients (i.e., road traffic accidents, fall down from height by accident or suicide, physical abuse), patients who were not personally observed by the authors, and patients meeting the inclusion criteria that could not be enrolled to the study due to refusal of informed consent were excluded from the study.
Statistical analysis
All data were analyzed using Statistical Package for Social Sciences (SPSS) for Windows 20 (SPSS Inc., Chicago, IL, USA) statistical software. The normality of data distribution was evaluated with Kolmogorov-Smirnov and Shapiro-Wilk tests. Variables showing normal distribution were expressed as mean ± standard deviation (SD), and variables not showing normal distribution were expressed as median (minimum-maximum). Mann Whitney U test was used for comparison of groups in variables not showing normal distribution. Pearson χ2-test was used for comparison of categorical variables. Multivariate analysis of independent variables was performed using logistic regression analysis. The level of statistical significance was set at p<0.05.
Results
A total of 498 patients were admitted to PICU during the study period, and 150 patients met the inclusion criteria. The male to female ratio was 1.05, with 77 (50.3%) male and 73 (47.7%) female patients. The mean age of the patients was 5.2±5.6 (min 0.1, max 17, median 2.4) years. Demographic data are summarized in Table 1. Neurological diseases were the most frequent admission diagnosis (n=54, 36%), followed by metabolic disorders (n=19, 12.7%) and pulmonary diseases (n=18, 12%). According to our results, the reason for RH was commonly neurological (n=3), after that hematologic diseases (1%) and carbon monoxide intoxication (0.7%). The prevalence of RH was 3.33% (5/150). The effect of endotracheal/tracheostomy tube, arrhythmia, sepsis, systolic and diastolic blood pressure values (mm Hg), presence of seizures, pneumothorax, and coagulation disorders/coagulopathy on RH prevalence was 2.7%.
However, the study demonstrated that the risk factors of hyperkalemia (odds ratio 17.8; 95% confidence interval (CI)1.3-240, p<0.05), cardiopulmonary resuscitation (odds ratio 10.1; 95% CI 1.1-93.5, p<0.05), and right heart failure (p<0.05) were significantly associated with RH (Table 1). Details on patients with RH and ophthalmologic findings are shown in Tables 2, 3 and 4.
Discussion
Although there is adequate literature on trends in determining RH, there are few studies in pediatric population and the information in developing world about this subject is uncertain. The majority of studies start from a suggestion of child abuse and the incidence of RH in pediatric population when newborns are excluded is ambiguous. Therefore, this study was performed to define the prevalence, distribution, and extent of RH in critically ill children admitted to PICU excluding newborns and patients with trauma. There were 77 (50.3%) male and 73 (47.7%) female patients. These findings were similar to the literature12,15. According to our results, the mean age of patients was 5.2±5.6 (median: 2.4) years and we found that children aged less than 2 years predominated (n=79, 51.7%). Consistent with our findings, median age was 1.5 years12 and 2 years10 in two studies. It is complicated to compare age distribution among studies because inclusion criteria are various in each study. RH is a frequent finding in child abuse, especially in the shaken baby syndrome which is a serious problem without obvious sign of physical injury9,15. Most studies report that RH is not indicative of a preceding trauma and must not be misinterpreted as child abuse because it may also be associated with CPR, accidental trauma, and various diseases1,11,16,17.
Retinal hemorrhage is common in non-accidental head injury, particularly in the shaken baby syndrome, with the incidence of 38%-100%1,18. In contrast, no RH was found in a study of 108 infants with apparent life-threatening events19. We report a 3.33% prevalence of the presence of any RH in critically ill children, with no severe multilayered RHs, conversely to those reported as 15.1% in critically ill pediatric population12,20. Gayle et al. state that RH occurs in 10% of children presenting no evidence of trauma with any other risk factor for hemorrhage. Some studies suggest that the retinal venous system in infancy is more vulnerable to increased intrathoracic pressure1,11. Consistent with our findings, retinal bleeding was mostly mild, not extensive (75%). In addition, RH in PICU was not usually severe, multilayered or extensive. The rate of severe bilateral multilayered RH was 3.8% and was completely in association with severe trauma, sepsis, coagulopathy or a combination of these factors12,20. A review has described the association of severe bilateral extensive RH with abusive head trauma7. Binenbaum et al. prospectively examined fundus of children undergoing lumbar puncture as part of their clinical care and found that only a small proportion of children with non-traumatic elevated intracranial pressure (16%) had RH21. In our study, the reason for no case of severe RH might be that our study excluded patients with abusive head trauma, penetrating eye trauma, and all other trauma patients. A literature review suggests that the severity of RH is more valuable in predicting abuse than the presence of RH8.
In our study, CPR and heart failure were significantly associated with RH. In some studies, RH has been reported as a sequel of CPR1,10. Chest compression from any cause raises intrathoracic pressure and this induces RH by directly increasing retinal venous pressure. In addition, simultaneous chest inflation with compression increases the right atrial pressure. Therefore, CPR produces salient pulsatile venous hypertension of the head and neck. Increased intracranial pressure by compression has been shown to cause retinal venous hypertension and RH1,10. This might be the reason that in our study heart failure was significantly associated with RH besides CPR.
Hyperkalemia was stated to be a precipitating event for RH in cardiopulmonary arrests requiring chest compression4. In our study, hyperkalemia was significantly associated with RH. Studies are limited regarding effects of hyperkalemia on RH and this was a key finding drawing attraction to electrolyte imbalances that might cause or precipitate RH. Further studies are necessary on this issue as soon as possible.
Like most studies, we found that there was no detectable impact of age, age group, gender, admission diagnosis, respiratory failure, requirement of mechanical ventilation, presence of an endotracheal/tracheostomy tube, arrhythmia, sepsis, hypo/hypertension, seizures, pneumothorax, and coagulopathy on the prevalence of RH. This was all highly correlated with the literature1,22,23. However, some studies stated that the presence of coagulation disorders/coagulopathy increased the prevalence of RH as expected1,3,12. The reason might be that there were only 23 (15%) patients presenting with coagulation disorders.
It is uncommon that a child with shaking impact injury would attend initially an ophthalmologist and our experience suggests that most pediatric intensivists/emergency physicians have difficulty in evaluating the fundus of a child. Considering diagnostic importance of RH, it seems important for a pediatric intensivist/emergency physician to learn through examination of the fundus and pupil dilatation about visualization of retina in a child when necessary. If there is suspicion of RH, an ophthalmologist’s opinion should be required for confirmation and documentation. The challenges ophthalmology faces today must be overcome by developing long-term strategies connected with remembrance of healthcare system values and diagnostic procedures as physicians24-26. Any child found to have RH might have an underlying medical condition that requires further investigations.
To our knowledge, our research is an important prospective study estimating the prevalence of RH in a population of critically ill children without any kind of trauma. By this means, we hope to decrease the risk of RH underdiagnosing. The strength of our study was that by excluding patients with trauma we provided an outcome more likely to approach the baseline incidence of RH in all sick children. Limitations of this study may include the absence of a control group of healthy children, and unfortunately, the patients meeting the inclusion criteria could not be enrolled to the study due to refusal of informed consent. Additional studies with larger samples should be carried out to warrant these results.
Conclusion
We report a 3.33% prevalence of the presence of any RH in critically ill children, mostly mild and no severe bilateral multilayered RH. The risk factors of hyperkalemia, cardiopulmonary resuscitation and heart failure were significantly associated with RH. In addition, a study focused on the effects of RH on visual acuity and visual outcome in the future will be helpful for understanding the long-term impact of RH.
References
1. Gayle MO, Kissoon N, Hered RW, Harwood-Nuss A. Retinal hemorrhage in the young child: a review of etiology, predisposed conditions, and clinical implications. J Emerg Med. 1995;13(2):233-9. doi:10.1016/0736-4679(94)00145-6
2. Gan NY, Lam WC. Retinal detachments in the pediatric population. Taiwan J Ophthalmol. 2018;8(4):222-36. doi: 10.4103/tjo.tjo_104_18
3. Aryan HE, Ghosheh FR, Jandial R, Levy ML. Retinal hemorrhage and pediatric brain injury: etiology and review of the literature. J Clin Neurosci. 2005;12(6):624-31. doi:10.1016/j.jocn.2005.05.005
4.Levin AV. Retinal hemorrhage: science versus speculation. J AAPOS. 2016;20(2):93-5. doi: 10.1016/j.jaapos.2016.03.001
5. Wegner A, Schmidt KG. The incidence of retinal haemorrhages and their implications. Br J Ophthalmol. 2013;97(8):1082-3.http://dx.doi.org/10.1136/bjophthalmol-2013-303659
6. Watts P, Maguire S, Kwok T, Talabani B, Mann M, Wiener J. Newborn retinal hemorrhages: a systematic review. J AAPOS. 2013;17(1):70-8. doi:10.1016/j.jaapos.2012.07.012
7. Maguire SA, Watts P, Shaw A, Holden S, Taylor R, Watkins WJ. Retinal haemorrhages and related findings in abusive and non-abusive head trauma: a systematic review. Eye. 2013;27(1):28-36. doi: 10.1038/eye.2012.213
8. Bhardwaj G, Chowdhury V, Jacobs MB, Moran KT, Martin FJ, Coroneo MT. A systematic review of the diagnostic accuracy of ocular signs in pediatric abusive head trauma. Ophthalmology. 2010;117(5):983-92. doi:10.1016/j.ophtha.2009.09.040
9. Morad Y, Kim YM, Armstrong DC, Huyer D, Mian M, Levin AV. Correlation between retinal abnormalities and intracranial abnormalities in the shaken baby syndrome. Am J Ophthalmol. 2002;134(3):354-9. doi:10.1016/s0002-9394(02)01628-8
10. Goetting MG, Sowa B. Retinal hemorrhage after cardiopulmonary resuscitation in children: an etiologic reevaluation. Pediatrics. 1990;85(4):585-8. PMID: 2314970
11. Levinson JD, Pasquale MA, Lambert SR. Diffuse bilateral retinal hemorrhages in an infant with a coagulopathy and prolonged cardiopulmonary resuscitation. J AAPOS. 2016;20(2):166-8. doi: 10.1016/j.jaapos.2015.11.003
12. Agrawal S, Peters MJ, Adams GG, Pierce CM. Prevalence of retinal hemorrhages in critically ill children. Pediatrics. 2012;129(6): e1388-96. doi:10.1542/peds.2011-2772
13. Levi M, Toh CH, Thachil J, Watson HG. Guidelines for the diagnosis and management of disseminated intravascular coagulation. British Committee for Standards in Haematology. Br J Haematol. 2009;145(1):24-33. doi:10.1542/peds.2011-2772
14. Kliegman RM, Stanton B, St Geme JW III, Schor NF, Behrman RE, editors. Nelson Textbook of Pediatrics. 21st edn. Philadelphia, PA, USA: Elsevier; 2019.
15. Kivlin JD. A 12-year ophthalmologic experience with the shaken baby syndrome at a regional children’s hospital.Trans Am Ophthalmol Soc. 1999;97:545-81. PMID: 10703141
16. Kaur B, Taylor D. Fundus hemorrhages in infancy. Surv Ophthalmol. 1992;37(1):1-17.https://doi.org/10.1016/0039-6257(92)90002-B
17. Riffenburgh RS, Sathyavagiswaran L. Ocular findings at autopsy of child abuse victims. Ophthalmology. 1991;98(10):1519-24. doi:10.1016/s0161-6420(91)32095-5
18. Duhaime AC, Alario AJ, Lewander WJ, Schut L, Sutton LN, Seidl TS. Head injury in very young children: mechanisms, injury types, and ophthalmologic findings in 100 hospitalized patients younger than 2 years of age. Pediatrics. 1992;90:179-85. PMID: 1641278
19. Curcoy AI, Trenchs V, Morales M, Serra A, Pou J. Retinal hemorrhages and apparent life-threatening events. Pediatr Emerg Care. 2010;26(2):118-20. doi:10.1097/PEC.0b013e3181cfdb6b
20. Adams GG, Agrawal S, Sekhri R, Peters MJ, Pierce CM. Appearance and location of retinal haemorrhages in critically ill children. Br J Ophthalmol. 2013;97(9):1138-42.http://dx.doi.org/10.1136/bjophthalmol-2012-302429
21. Binenbaum G, Rogers DL, Forbes BJ, Levin AV, Clark SA, Christian CW, et al. Patterns of retinal hemorrhage associated with increased intracranial pressure in children. Pediatrics. 2013;132(2):e430-4. doi:10.1136/bjophthalmol-2012-302429
22. Mei-Zahav M, Uziel Y, Raz J, Ginot N, Wolach B, Fainmesser P. Convulsions and retinal haemorrhage: should we look further? Arch Dis Child. 2002;86(5):334-5. http://dx.doi.org/10.1136/adc.86.5.334
23. Tyagi AK, Scotcher S, Kozeis N, Willshaw HE. Can convulsions alone cause retinal haemorrhages in infants?Br J Ophthalmol. 1998;82(6):659-60.http://dx.doi.org/10.1136/bjo.82.6.659
24. Grewe R. Ophthalmology in the European Union at the beginning of the 21st century and its opening towards Eastern Europe. Graefes Arch Clin Exp Ophthalmol. 2000;238(10):812.
25. Kaštelan S, Antunica AG, Rabatić JS, Gotovac M, Orešković D, Kasun B. Traumatic optic neuropathy – case report with discussion on diagnostic procedures and therapy. Acta Clin Croat. 2018;57(1):166. doi: 10.20471/acc.2018.57.01.22
26. Zavoreo I, Jurašić MJ, Lisak M, Tomas AJ, Zrinščak O, Bašić Kes V. Recurrent atypical optic neuritis as the leading sign of Fabry disease. Acta Clin Croat. 2019;58(3):550. doi: 10.20471/acc.2019.58.03.22
Sažetak
Krvarenje mrežnice u kritično bolesne djece – prospektivno istraživanje
Tülin Köksal, Şanlıay Şahin, Esra Dağ Şeker, Ganime Ayar, Mutlu Uysal Yazıcı i Alkım Öden Akman
Cilj ovoga opservacijskog istraživanja bio je utvrditi učestalost, raspodjelu i opseg krvarenja mrežnice (KM) u kritično bolesne djece, njihove demografske značajke, klinička obilježja i ishod bolesnika s KM, isključujući bolesnike s traumom. Također smo nastojali opisati utjecaj dobi, spola, dijagnoza kod prijma, čimbenika rizika poput repiracijskog zatajenja, potrebe mehaničke ventilacije, prisutnosti endotrahealne/traheostomske cijevi, kardiopulmonalne reanimacije, aritimije, hiperkalijemije, pneumotoraksa, sepse, srčanog zastoja, hipo/hipertenzije, prisutnosti napadaja, prisutnosti koagulacijskih poremećaja na tu raspodjelu kao sekundardni rezultat. Podatci su prospektivno prikupljani uzastopno od travnja 2013. do travnja 2014. godine. Za vrijeme istraživanja utvrđeno je ukupno 150 bolesnika koji su odgovarali uključnim kriterijima. Omjer muških i ženskih bolesnika bio je 1,05. Nije bilo uočljivog utjecaja dobi, dobne skupine, spola, dijagnoze kod prijma, respiracijskog zatajenja, potrebe mehaničke ventilacije, prisutnosti endotrahealne/traheostomske cijevi, aritmije, sepse, hipo/hipertenzije, prisutnosti napadaja, pneumotoraksa, prisutnosti koagulacijskih poremećaja/koagulopatije na učestalost KM. Istraživanje je pokazalo da su rizični čimbenici poput hiperkalijemije, kardiopulmonalna reanimacija i srčani zastoj značajno povezani s KM. Svako dijete koje ima KM moglo bi patiti od nekog osnovnog medicinskog stanja koje zahtijeva daljnje pretrage. Usto, buduća studija usredotočena na učinke KM na vidnu oštrinu i ishod vida pomoći će u razumijevanju dugoročnog utjecaja KM.
ključne riječi
Djeca, kritično bolesna; Pedijatrijska jedinica intenzivnog liječenja; Krvarenje mrežnice
table 1. Demographic characteristics of study patients
table 2. Risk factors for retinal hemorrhage in children (N=150)
* Odds ratio and 95% confidence intervals
table 3. Details on patients with retinal hemorrhage and ophthalmologic findings
table 4. Details on patients with retinal hemorrhage and ophthalmologic findings in critically ill children
MV = mechanical ventilation; ETT = endotracheal tube; CPR = cardiopulmonary resuscitation