Premature infants endure numerous painful procedures during their stay in the neonatal intensive care unit (NICU); one of these procedures is weekly (or biweekly) ophthalmic examinations. To directly visualize the anatomy of the eye, ophthalmologists rely on tools that keep the infants' eyes open while manipulating them from side to side. These examinations are performed to monitor for the development of retinopathy of prematurity (ROP), a disease known to cause infant blindness, until maturation of the retinal vessels occurs, or there is resolution of the already diagnosed disease. According to the American Academy of Pediatrics, the frequency of these examinations is crucial because early detection and treatment of ROP is necessary for improved visual outcomes.1 Ophthalmic examinations may also be indicated in neonates with congenital defects, abnormal fundoscopic findings, or suspected infections. Although important, these examinations are painful and stressful for all infants. In one study,2 it was found that infants had clinical deterioration during ROP examinations, which included increased blood pressure, increased heart rate, and more frequent oxygen desaturation episodes. In addition to these physiological changes, neonatal pain and stress are displayed through behavioral cues, such as hiccupping, sneezing, yawning, grimacing, mottling, and hyperextension of extremities.3 Research has shown that repetitive painful procedures, such as ROP examinations, have long-term consequences, which may include pain sensitivity, behavioral problems, and learning disabilities.4 Therefore, it is important to provide supportive interventions to all infants during and following ophthalmic examinations to minimize the associated pain and stress.
EVIDENCE FOR INTERVENTION
Supportive interventions associated with improved infant outcomes include nutritive sucking with oral sucrose or human milk, topical anesthetic eye drops, and facilitated tucking or containment. Oral glucose solutions5 and orally administered human milk6 have been shown to significantly reduce pain scores during ophthalmic examinations when compared with standard care. Administration of topical anesthetics prior to ophthalmic examinations was demonstrated to be effective in reducing pain scores when paired with other supportive interventions.4 Facilitated tucking and containment during painful procedures have been associated with significantly lower pain scores compared with standard care7; furthermore, significant reductions in pain were demonstrated when facilitated tucking was paired with the administration of oral glucose solutions during ophthalmic examinations.8
PROBLEM AND PURPOSE STATEMENT
While some NICUs develop their own, unit-based ophthalmic examination protocol, there are currently no mandatory practice guidelines in the United States to include supportive interventions during examinations, despite the evidence to support their use.9 The purpose of this quality improvement project was to implement evidence-based, supportive interventions during ophthalmic examinations in premature infants and evaluate the impact on pain, oxygen saturation, heart rate, number of bradycardic events, and number of neonatal stress cues.
METHODS
Setting and Sample
This quality improvement project was conducted in an 80-bed, level III NICU in Pittsburgh, Pennsylvania. In our NICU, there are 10 to 20 eye examinations per week on average. A convenience sample of infants was used for this project, and all infants receiving ophthalmic examinations in the NICU were eligible for inclusion in the project unless they received a paralytic, sedative, or narcotic in the previous 24 hours.
Interventions
The supportive interventions provided to all infants in this project were (1) at least 5 minutes of undisturbed rest prior to the examination, (2) anesthetic eye drop administration immediately prior to the start of the examination as per current hospital policy, (3) hand containment or swaddling provided 5 minutes before, during, and 5 minutes after the examination, and (4) the offering of a pacifier dipped in a 24% sucrose solution or human milk 5 minutes before, during, and 5 minutes after the examination, if medically appropriate. These interventions were provided to each infant by either the project leader (registered nurse) or the developmental specialist to ensure quality control and consistent intervention delivery throughout the duration of the project.
Design
A preintervention and intervention group design was used for this quality improvement project. The duration of the preintervention and intervention periods was 1 month each. Demographic and clinical information was collected on all infants in the preintervention and intervention groups. Outcome measures were assessed on infants who received standard care during ophthalmic examinations and compared with the outcome measures collected on infants who received the supportive interventions during ophthalmic examinations. Standard care during ophthalmic examinations at this facility included only anesthetic eye drop administration prior to the examination and head stabilization during the examination to minimize movement. The preintervention and intervention group outcome measures were assessed at 3 different time points during the examination process, which included (1) 5 minutes before the examination, (2) during the examination, and (3) 5 minutes after the examination.
Outcome Measures
The primary outcomes assessed in this project were pain and oxygen saturation measured in the preintervention and intervention groups. Pain and oxygen saturation were assessed 5 minutes before each examination, during each examination, and 5 minutes after each examination. Pain was assessed using the Premature Infant Pain Profile-Revised (PIPP-R) tool, which is a standardized, multidimensional pain scale with established validity and reliability in assessing procedural pain in term and preterm infants.10 PIPP-R scores were assigned at the designated time points by a consistent project team member, who was trained to use the tool. Infant oxygen saturation was assessed via bedside cardiopulmonary monitors, and the lowest percentage oxygen saturation value observed was recorded at each time point.
Secondary outcomes assessed in this project were heart rate, number of bradycardic events, and number of neonatal stress cues. These outcomes were also measured in the preintervention and intervention groups and were assessed 5 minutes before each examination, during each examination, and 5 minutes after each examination. Heart rate in beats per minute was assessed via bedside cardiopulmonary monitor, and the maximum heart rate observed was recorded for each time point. Number of observed bradycardic events (<80 beats per minute) was also documented when present. Neonatal stress cues were assessed by observation of the infant and the presence of each cue was recorded by a consistent project team member at each time point when observed. The stress cues that were assessed included hiccupping, sneezing, yawning, grimacing, mottling, and hyperextension of extremities.
Data Collection Procedures
Demographic and clinical data (gestational age at birth, corrected age at the time of the examination, gender, weight in grams, and current respiratory support [none, nasal cannula, bubble continuous positive airway pressure or continuous positive airway pressure, and endotracheal tube]) were collected on all infants included in the project on the day of their examination via electronic medical records and information provided by the bedside nurse. Outcome measurements (PIPP-R score, lowest oxygen saturation value, maximum heart rate, number of bradycardic events, and number of neonatal stress cues) were manually documented at the patient's bedside by a consistent project team member via standardized documentation sheets (see Supplemental Digital Contents 1 and 2, available at http://links.lww.com/ANC/A66 and http://links.lww.com/ANC/A67, respectively). The standardized documentation sheets excluded patient identifiers and were entered into an Excel spreadsheet for data management.
Statistical Analysis
The preintervention and intervention groups were assessed for similarity through comparison of the demographic and clinical information collected for each case. [chi]2 tests were used to compare gender and current respiratory support between groups, and independent-sample t tests were used to compare means of gestational age, corrected age, and weight between groups. The means of each outcome measure at each time point were compared between the preintervention and intervention groups using independent-sample t tests. All analyses were conducted using SPSS Statistics v. 26.0 (IBM Corporation, Armonk, New York) with a significance level set to .05. Cohen's d values with 95% confidence intervals were calculated to determine the effect size of the supportive interventions on the outcomes at each time point.
Ethical Considerations
This project was submitted to and approved by the Quality Review Committee of the health system to which the NICU belonged and the Newborn Quality Assurance Committee for the project site. Institutional review board approval was not required for this project.
RESULTS
Patient Characteristics
The sample included 60 infants, with 30 in the pre-intervention group and 30 in the intervention group. There were no significant differences among the 2 groups in gestational age, corrected age, gender, weight, or respiratory support required at the time of the examination (Table 1).
Outcome Results
Table 2 displays the results for the comparison of outcome measures between the preintervention and intervention groups. There was a trend for the average pain score in the intervention group to be lower than the average pain score in the preintervention group before the examination (P = .052). The average pain scores in the intervention group were significantly lower than the average pain scores in the preintervention group during (P < .001) and after (P = .001) the examination. There were no significant group differences in the mean oxygen saturation values and mean heart rates before, during, and after the examination. The mean number of bradycardic events was significantly lower in the intervention group than in the preintervention group during (P = .004) and after (P = .039) the examination. Lastly, the mean number of neonatal stress cues was significantly lower in the intervention group before, during, and after the examination when compared to the preintervention group (P = .002, P < .001, and P < .001, respectively).
Cohen's d values were also computed for all outcome measures at the different time points. Very large effect sizes (d +/- 1.00) favoring the supportive interventions were found for pain scores during the examination and the number of neonatal stress cues during and after the examination. Large effect sizes (d +/- 0.80) favoring the supportive interventions were found for pain scores after the examination and neonatal stress cues before the examination. A medium effect size (d +/- 0.50) in favor of the supportive interventions was found for pain scores before the examination. Small effect sizes (d +/- 0.20) in favor of the supportive interventions were found for oxygen saturation during the examination and the number of bradycardic events during and after the examination.
DISCUSSION
This project demonstrated that supportive interventions during infant ophthalmic examinations, which included rest and anesthetic eye drops prior to the examination, containment or swaddling, and a pacifier dipped in sucrose or human milk, significantly reduced pain, number of bradycardic events, and neonatal stress cues associated with examinations. The effect of the supportive interventions on these outcomes was small to very large depending on the outcome and time point. However, the interventions did not affect lowest oxygen saturation values or maximum heart rates associated with examinations. Possible reasons for nonsignificant group differences in lowest oxygen saturation values were: differences in underlying disease processes among patients (pneumonia, respiratory distress syndrome, etc), inaccurately displayed oxygenation values on the bedside monitor (excessive movement of the SpO2 probe minimizes accuracy of the results), and differences in respiratory support and supplemental oxygen among patients. Supplemental oxygen for infants was not recorded as part of this project, but may be useful in future projects to better evaluate changes in oxygenation related to the intervention. Potential reasons for the difference between observed and anticipated outcomes for maximum heart rate include differences in baseline heart rates among patients, differences in underlying disease processes (anemia, sepsis), and activities prior to the examination that caused excessive agitation (care, labs, etc), which was not completely resolved in the 5-minute rest period prior to the examinations.
Previous research has supported the use of supportive interventions in reducing pain and stress cues associated with ophthalmic examinations in premature infants.4-8 The interventions and protocols used in these studies were different based on the specific needs of the units in which they were implemented. Additionally, there was some variation among studies in the supportive interventions, such as the sweetener administered on pacifiers (glucose, sucrose, or human milk) and the containment method used (facilitated tucking or swaddling). Currently, there are no data to support that one specific sweetener is more effective over the other, but there is evidence to support improved outcomes with the use of any sweetener instead of offering nothing at all. Similarly, no one specific containment method appears to be superior over the other, but improved outcomes are apparent when any containment method is used compared with standard care.4 Overall, the implementation of supportive interventions as part of a unit-based project is a low-cost, effective way to improve outcomes in premature infants, and successful implementation of a protocol is more likely to be accomplished when interventions are chosen based on the specific needs of the unit and processes for implementation are designed based on input from important stakeholders, such as the ophthalmologists, nurses, and other care team members.
The first identified barrier to implementing a protocol with supportive interventions at this facility was staff resistance to change. At the facility in which this project took place, there were dedicated nurses that rounded with the ophthalmologist during eye examinations. While some were very receptive to the idea of the new protocol interventions, others were resistant due to the time constraints and additional work required. For example, if a nurse was assigned to assist with eye examinations for the morning, the nurse was usually assigned for 2 hours before having to take a patient assignment. To address these concerns, changes were made to the project to allow for a quicker transition time between patients (infants were given 5 minutes of intervention following the examination instead of >=10) and a designated project team member would get the next patient ready while the previous patient was recovering. Another staff issue encountered was the overall lack of involvement of the bedside nurse in the eye examination process. Because this unit had a designated nurse assigned to assist with examinations, the bedside nurses were uninvolved, which led to the patients not being ready or at rest prior to the examination. To address this barrier, the project team members rounded with bedside nurses on the morning of examination day to remind them that their patient had an examination, to ensure the infant was provided undisturbed rest prior to the examination by delaying routine care until after the examination, and to have oral sucrose or human milk ready at the bedside for administration with a pacifier. This rounding strategy helped to communicate the needs of the ophthalmic examination team to ensure interventions were delivered in a timely manner and according to the guidelines of the project.
The second identified barrier encountered during protocol implementation was that oral sucrose and human milk were not readily available for the examinations. This was due to the following factors: (1) many staff were unaware of when they needed to have the supplies ready for the examination, (2) a new human milk scanning process was rolled out at the same time the project started, and (3) oral sucrose was considered a medication at this facility, meaning patients had to have an order in their medication-administration record for scanning. Since these scanning processes increased the time spent getting the patient ready for examinations, resistance from the eye examination nurses increased. To address this barrier, the project team members asked the bedside nurses to have the solutions scanned and ready at the bedside to reduce the time the ROP nurse had to spend in each patient's room. Additionally, the project team members worked with the medical team and the information technology staff to get the eye examination order set adjusted to include sweet-ease solution as an ordered medication so that all infants had the option to have sweet-ease for the examination if human milk was not available.
Limitations
One limitation of this project was the small sample size due to time constraints for project completion. A larger sample could show additional significant results. Additionally, some infants in this project did not have SpO2 monitors (5 in the preintervention group and 10 in the intervention group); therefore, oxygen saturation data were missing on these infants. Another limitation of this project was the timing of the ophthalmic examinations. Each infant had an examination at a different time of the morning; some examinations occurred before care times, others interrupted care time, and some occurred immediately following care time or a feeding. Because the timing of the examination was such a large threat to the project's internal validity, the project team members ensured that at least 5 minutes of rest was given to each infant prior to the examination so that outside stressors were minimized prior to the start of the examinations. Five minutes was the chosen amount of time used in this project for the convenience of the ophthalmic examination team, but more undisturbed rest prior to each examination may increase the validity of the outcomes due to the intervention.
Implications for Practice
This quality improvement project provides NICUs with a template for providing supportive interventions during ophthalmic examinations, which have been demonstrated to improve patient outcomes. The development of similar protocols should address the specific needs of each NICU. For example, based on the ophthalmic examination process, bedside nurses may administer interventions or this task may be delegated to other team members, such as an ROP nurse, developmental specialist, or ancillary staff member. Key stakeholders should be involved in the planning of the project and in the implementation of the new protocol to gain a better understanding of current procedures and potential barriers. Staff education is also essential to the success of a new protocol so that there is an understanding of the importance of the intervention being provided as well as the research evidence to support its use. Furthermore, providing education on these interventions may empower nurses to provide similar interventions in other painful procedures, such as intravenous needle insertions, blood draws, and suctioning.
CONCLUSIONS
In this quality improvement project, the implementation of a unit-based protocol, which included supportive interventions during ophthalmic examinations in premature infants, was demonstrated to be effective in reducing pain, bradycardic events, and neonatal stress cues associated with the examinations. This project validated that incorporating these interventions into a unit-based protocol for ophthalmic examinations is an achievable and cost-effective way to improve patient outcomes.
References