Physiological responses to retinopathy of prematurity screening: indirect ophthalmoscopy versus ultra-widefield retinal imaging

Abstract

Background/aims: Retinopathy of prematurity (ROP) screening is vital for early disease detection in very premature infants but can cause physiological instability. This study compares the physiological response to binocular indirect ophthalmoscopy (BIO) with indentation and non-contact ultra-widefield (UWF) retinal imaging in non-ventilated neonates. The impact of the Dandle WRAP, a specialised swaddling aid, on UWF imaging was also assessed.

Methods: This retrospective study included 86 ROP screening events in 66 non-ventilated infants aged 35.3 weeks (range 30.6-44.6). Vital signs were continuously recorded, evaluating immediate (within 15 min) and longer-term (within 12 h) physiological responses.

Results: ROP screening significantly increased heart and respiratory rates and decreased oxygen saturation within 15 min of screening. No significant differences in physiological responses were found between BIO and UWF imaging, although there was a trend towards lower maximum heart rate with UWF imaging. The Dandle WRAP did not significantly alter physiological responses but improved the ease and speed of UWF imaging.

Conclusion: UWF imaging does not increase physiological instability compared to BIO in non-ventilated infants. Specialised swaddling aids may facilitate the imaging procedure.

Impact: ROP screening can be distressing for premature infants and induce physiological instability during and after the examination. We deployed non-contact ultra-widefield retinal imaging as the default method of ROP screening and show that it induces comparable physiological responses as traditional indirect ophthalmoscopy in non-ventilated babies. Dandle WRAP swaddling facilitated handling and speed of retinal imaging. The study demonstrates that imaging-based ROP screening is safe and efficacious in non-ventilated neonates, and continuous multimodal physiological recordings can provide detailed assessment of the effects of procedures and medications.

Fig. 1. Screening methods and study flowchart.

a–d Illustrative examples of ROP screening techniques and use of the Dandle WRAP. Parents provided written consent for the use of images. a An infant undergoing traditional ROP screening by binocular indirect ophthalmoscopy (BIO) with indentation. b An unswaddled infant undergoing ultrawidefield (UWF) retinal imaging in the ‘flying baby’ holding position. Note that the infant’s arms needed to be stabilised by an assistant while a second assistant captured the images. c An infant fully swaddled with a Dandle WRAP whilst resting in their incubator. d An infant swaddled with a Dandle WRAP undergoing UWF imaging. Note that the legs can be removed from the Dandle Wrap to improve stability and safety in the flying baby position. As the infant’s arms are securely swaddled, only one assistant was needed to capture the images. e Study flowchart. For Study 3, infants were divided into groups according to whether they were swaddled using the Dandle WRAP (DW) or unswaddled (control group) during UWF imaging. Note that infants in the control group were swaddled with a conventional blanket whilst in their cot/incubator but were not swaddled during UWF imaging as it would be unsafe to perform the flying baby hold with a full swaddle.

Fig. 2. Comparison of physiological changes following ROP screening by UWF imaging versus BIO.

a–f Changes in the 15 min following the start of ROP screening. Changes in a heart rate (HR), b oxygen saturation (SpO₂) and c respiratory rate (RR) in response to ROP screening. Solid lines indicate the group mean and shaded areas represent standard deviation. Red represents the infants who were screened using Binocular indirect ophthalmoscopy (BIO). Black represents the infants who were screened using UWF (ultra-widefield) imaging. Individual infant traces are baseline corrected by subtracting their pre-procedure mean. Grey vertical line indicates the start of ROP screening. d–f Comparison of physiological metrics between BIO and UWF imaging groups measured in the 15 min following the start of ROP screening shown using violin plots where the width of the plot represents the number of infants at that value. Horizontal dotted lines indicate mean and interquartile range. d Metrics (average and maximum) used to assess HR changes, in beats per min. e Metrics used to assess SpO₂ changes. f Metrics used to assess RR in breaths per min. g–l Changes in the 12 h following the screening. j Number of bradycardia (<100 bpm for at least 15 s) and number of tachycardia (>200 bpm for at least 15 s), k number of desaturations (<80% for at least 10 s), and l number of apnoeas (pauses in breathing of at least 20 s) in the 12 h pre and post screening.

Fig. 3. Comparison of infant physiology in response to ROP screening with different swaddling interventions.

Changes in a heart rate, b oxygen saturation and c respiratory rate in response to ROP screening. Solid line indicates the group mean and shaded areas the standard deviation. Blue represents Control group infants, orange represents infants who were swaddled using the dandle WRAP (DW). Individual infant traces are baseline corrected by subtracting their pre-procedure mean. Grey vertical line indicates the start of ROP screening. All infants were screening using UWF imaging. d–f Comparison of physiological metrics between the two groups measured in the 15 min following the start of ROP screening shown using violin plots where the width of the plot represents the number of infants at that value. Horizontal dotted lines indicate mean and interquartile range.