Incidence of Retinopathy of Prematurity and Its Association with Oxygen Therapy in Preterm Low Birth Weight Babies

Address for correspondence: Dr. Somen Sur 67/E/2, GT Road (West), PO Mullickpara, PS Serampore, Dist. Hooghly, West Bengal, India, Pin: 712203. E-mail: somensur29@gmail.com Tel No; +91 9433428988 1Dr. AK Pal, MD-PGT, Department of Paediatrics, Vivekananda Institute of Medical Sciences, Kolkata, India, 2Dr. S Sur, MD-PGT, Department of Paediatrics, Vivekananda Institute of Medical Sciences, Kolkata, India, 3Dr. J Pal, PGT, Department of Preventive and Social Medicine, Institute of Hygiene and Public Health, Kolkata, India, 4Dr. AK Gupta, Professor and Head of the Department, Department of Paediatrics, Vivekananda Institute of Medical Sciences, Kolkata, India. Abstract


Introduction
A ssocia on between oxygen and re nopathy of prematurity (ROP) has been known for a long me.Tin and colleagues reported a decrease incidence of ROP in whom oxygen is maintained between 70%-90% versus 88%-98% 1 .Aston and others 2 had suggested that the suscep bility of the growing vessels in the re na to hyperoxia probably related to unique anatomic and developmental rela onship between the re nal and choroidal circula on 3 .Choroidal vessels supply the outer half of the re na by diff usion.They are more permeable, and have a high venous PaO 2 , lack the ability of autoregula on in response to hyperoxia.PaO 2 level is raised across the thickness of the re na and re nal vessels respond by constric on.However a meta-analysis by Askie and Hendersonsmart 4 of 5 randomized control trials conducted in preterm infants in which ambient O 2 concentra on were targeted to achieve either a lower or higher blood O 2 range was unable to determine the op mal target range for maintaining blood PaO 2 levels in preterm infants.
The STOP-ROP study 5 was launched to see if development from prethreshold to threshold ROP can be reduced by hypoxia.Two groups were studied among babies with SaO 2 <94% in room air -the standard group with SaO 2 between 89%-95% and the high satura on group with SaO 2 between 96%-99%.Although a tendency for a be er outcome regarding the eye was found in low satura on group, these babies had signifi cantly more lung problems and also needed longer me on oxygen and more days in hospital.Several studies that looked at the eff ect of gradual versus abrupt weaning of oxygen on the incidence of ROP were unable to fi nd any diff erence between the two 6 .
In this study preterm babies having gesta onal age ≤35 wks were included.Screening was done at 4 wks of chronological age or at 31 week of gesta onal age (whichever was later) by indirect ophthalmoscope to evaluate the incidence and risk factors of ROP in preterm and/or low birth weight neonate and to associate incidence with oxygen therapy.

Material and Methods
This study was done at the Vivekananda Ins tute of Medical Sciences, Kolkata (Neonatal intensive care unit, Department of Paediatrics with help of department of Ophthalmology) over a one year period.It was a prospec ve observa onal study.Babies born with gesta onal age </= 35 weeks were included in this study.The number of neonates that fulfi lled the inclusion criteria was 50.
Binocular Indirect ophthalmoscope, Infant speculum and Kreissig scleral depressor were used for the tests.
Babies admi ed to the neonatal unit with gesta onal age of ≤35 weeks, were screened for ROP by an ophthalmologist by indirect ophthalmoscopy.Babies were examined at four weeks a er birth or 31 weeks gesta onal age, whichever was later.Mydriasis was achieved by using 1% Tropicamide thrice at intervals of 5 min followed by 1-2 drops of 2.5% or 5% phenylephrine twice at intervals of 5 min.Staging of ROP was done according to the Interna onal classifi ca on 7 .
Data obtained from this study were entered in Microso Excel and subsequently analysed in SPSS version 20.0.Descrip ve data was analysed as simple frequency, percentages, univariate analysis, odds ra o etc. and presented in tables and diagrams.For the en re sta s cal test applied, p value <0.05 had been considered to reject the null hypothesis.
Ethical clearance obtained from hospital and informed consents of parent were also obtained.In the babies screened, 24% had some evidence of ROP and 76% had no ROP detected.25% had stage 1, 41.7% had stage 2 and 33.3% had stage 3.In case of ROP in both eyes, the maximum stage of ROP was taken into considera on for analysis.The incidence of ROP 75%, 37.5%, 17.6% among ≤28 wk, 29-30 wk, 31-32 wk GA respec vely.The incidence of ROP was 83.3% among < 1000gm, 18.9% among >1000 gm.In Chi Square test birth weight was found to be a signifi cant risk factor for occurrence of ROP ( p value 0.001) Sixteen babies were ven lated, among them 11 developed ROP (91.6%).Eight babies received CPAP only & 1 developed ROP (8%).Among the babies receiving only free fl ow oxygen no one developed ROP.On analysis the Chi Square found that test ven la on was found to be a signifi cant risk factor (p-value <0.001) for development of ROP.

R e f
There was 72.6 mes greater risk of development of ROP in babies who had received oxygen therapy in form of ven la on than those who had received oxygen therapy in other forms (like CPAP/only free fl ow oxygen).76%) 50 (100%) Chi-Square=15.226,df=1,P=0.000,Among the babies who had received oxygen therapy of >150 hours, 60% had developed some form of ROP.Among the babies who had received oxygen therapy of <=150 hours, only 8.6% of them had developed ROP.Chi-square test shows that this diff erence is sta s cally signifi cant, that means dura on of oxygen therapy is a signifi cant risk factor for development of ROP.There is 16 mes higher risk of developing ROP if the dura on of oxygen therapy is >150 hours than in those where dura on of oxygen therapy is ≤150 hrs.This is sta s cally signifi cant as CI falls between 3.326 and 76.973.Above table shows that concentra on of oxygen delivery was a signifi cant risk factor for development of ROP.Above table shows that blood transfusion was not a sta s cally signifi cant risk factor for development of ROP in the present study.Above table shows that apnoea was not a sta s cally signifi cant risk factor for development of ROP in this study.

Discussion
Re nopathy of prematurity is a vasoprolifera ve disorder occurring predominantly in premature infants.Normal vasculature of the developing re na is interrupted due to some injury.A er a latent period, there is neovasculariza on.If this new vessel forma on is abnormal, it leads to progressive re nopathy, ul mately resul ng in re nal detachment and blindness 8 .ROP was fi rst reported in 1942 by Terry 9 .In spite of an intense search of a specifi c cause, it was not un l 1951 that a pediatrician in Melbourne, Australia provided the fi rst link between the use of oxygen and ROP 10 .Several report subsequently suggested that an inspired oxygen concentra on less than 40% was unlikely to lead to ROP 11,12 .The decline in use of greater than 40% supplemental oxygen for premature infant in 1950 resulted in the declining incidence of ROP 12,13 .The 1960s were associated with enormous advances in the fi eld of neonatology.It was then possible to resuscitate and ven late ny premature infants.Increased survival of these small infants together with the liberal use of oxygen led to resurgence of ROP 14 .
At present, par cularly in developed na ons, incidence of ROP is reduced by judicial use of oxygen therapy.In United States, around 10% of preterm babies are suff ering from blindness, which can be 20% or even higher in developing countries 15,16,17,18 .
Most common causes of ROP are premature birth and low birth weight 19,20 .Incidence of ROP increases with lower gesta onal age 21 .
Incidence of ROP in the present study was found to be 24% much lower than Rekha et al. 22 in 1996 but in more recent studies the incidence were more like this result eg Gupta et al. 23 2004 or Chowdhury et al. [24] 2009.Stage 2 (41.7%) was most common.In the present study, ROP was more prevalent (75%) in neonates born <28 wk gesta onal age comparable with study of Saeidi et al. 25 , which was signifi cant sta s cally (p 0.008), comparable with Hakeem et al., Karna et al. 21,26 .Prevalence of ROP was more in <1000 gram popula on (83.3%) comparable with Freeman et al., Kistner et al. 20,27 .There was also signifi cant associa on between ROP and low birth weight (p 0.001), mechanical ven la on (p <0.001), dura on of ven la on (p 0.000), comparable with Freeman et al. 27 , Shah et al. 28 , Mokhtari et al. 29 respec vely.Askie et al. through NeOProM study also concluded that high level of oxygena on increased the morbidity burden of re nopathy of prematurity 30 .In the present study, it is shown that increasing the concentra on of oxygen delivery carries signifi cant risk [p 0.000] for development of ROP.
Though Hassel et al. 31 and Banerjee et al. 32 had shown signifi cant associa on between blood transfusion and development of ROP, the present study could not show signifi cant associa on between them.Similarly, this study could not fi nd out any signifi cant risk associa on between apnoea and development of ROP, though according to Cha opadhyay et al. 33 and Aggarwal et al. 34 , apnoea carried signifi cant risk factor for development of ROP.Observa on of insignifi cant rela onship between apnoea and blood transfusion with ROP in the present study may probably be due to considera on of a small sample size.

Conclusion
The incidence of ROP was found around 24% in this study.The most important risk factors which were found to be signifi cantly predisposing to development of ROP included gesta onal age <=28 weeks, birth weight <1000 gram, ven la on as a type of oxygen therapy, use of increasing concentra on of oxygen delivery and oxygen therapy for >150 hours.
Thus, one should be cau ous regarding judicious use of oxygen therapy in premature babies par cularly during neonatal resuscita on (to use blender, to guide oxygen concentra on as per SPO 2 monitoring), to try to avoid prolonged ven la on, and to target SPO 2 85-95%.By ensuring these steps, one can cut down the incidence of ROP signifi cantly in future.
Further study with large sample size would be required in future to draw a more fi rm conclusion.

Table 1 :
Distribu on of study popula on according to the occurrence of ROP (n=50)

Table 2 :
Distribu on of the study popula on (n=12) depending on stage of ROP

Table 3 :
Distribu on of study popula on (n=50) depending on gesta onal age (GA)& incidence of ROP

Table 4 :
Table showing rela onship of ROP with gesta onal age of the babies In Chi Square test gesta onal age of the babies was found to be a signifi cant risk factor (p value -0.008) in occurrence of ROP.

Table 5 :
Table showing strength of associa on of ROP with gesta onal age of the babies There is 12.33 mes greater risk of development of ROP in babies with gesta onal age of <=28 weeks than babies with gesta onal age of >28weeks.

Table 6 :
Distribu on of popula on (n=50) depending on birth weight & incidence of ROP

Table 7 :
Table showing rela onship of ROP with birth weight of the babies

Table 8 :
Table showing strength of associa on of ROP with birth weight of the babies

Table 9 :
Table showing rela onship between type of oxygen therapy and ROP

Table 10 :
Table showing strength of associa on of development of ROP with type of oxygen therapy received by the babies

Table 11 :
Distribu on of study popula on (n=50) depending on dura on of oxygen therapy and incidence of ROP

Table 12 :
Table showing strength of associa on between oxygen therapy and incidence of ROP in popula on (n=50)

Table 13 :
Distribu on of study popula on (n=50) according to concentra on of oxygen delivered and incidence of ROP

Table 14 :
Distribu on of study popula on (n=50) according to requirement of blood transfusion (for anaemia) and incidence of ROP

Table 15 :
Distribu on of study popula on (n=50) according to occurrence of apnoea (once/ more) and incidence of ROP