INTRODUCTION AND PURPOSE
The traumatizing effect of having an infant born preterm and requiring a stay in the neonatal intensive care unit (NICU) has been well documented and is known to impact parent-infant interaction. Families in the NICU are often in a crisis state,1 and parents can develop stress disorders2 and experience increased cortisol levels,3 which may diminish learning abilities.3 Fortunately, family-based interventions designed to improve parent-infant interaction can reduce these symptoms in parents.4 Appropriate parental involvement is important to both the current and future health of the infant5 because developmental outcomes for infants born preterm rely on the infant's physical capabilities and on the parent-infant relationship.6 Sameroff and Chandler7 found that the quality of the parent-infant relationship and interaction, with the general care-giving environment, better predicted developmental outcome than traditional assessments.
McKim8 surveyed mothers of discharged high-risk preterm infants and found that the instructional areas of greatest need were interpreting infant cues and participating in routine daily care. Mothers reported they lacked confidence in caring for their infants and indicated that more information was needed in preparation for hospital discharge. Fostering parent-infant interaction and participation in care through instruction on infant behavior, handling techniques, and developmental activities are roles of the physical therapist in the NICU.9 , 10 Providing education on the right topic at the right time can be a challenge, because the NICU experience is unpredictable and variable.
The aim of this study was to promote improved parent-infant interaction by increasing parent competence in infant handling. The primary purpose of the study was to investigate the effectiveness of 3 different instructional methods on parent performance in the NICU during positioning and handling of their infants born preterm.
This study was a 3-way randomized blocked design with pre- and posttest analyses, using a forced-balance method for stratification and masked evaluator for follow-up. A computerized forced-balance randomization method was used to assign participants to 1 of the 3 instructional groups: (1) direct instruction and demonstration; (2) video instruction and demonstration; or (3) written-pictorial instruction by applying a priority weight to a stratification variable.11 To enhance balance between groups, a parameter was set to limit the differences in the numbers of participants between groups to 3. The randomization method used to balance the groups was based on 4 stratification criteria: (1) infant birth weight; (2) parent age; (3) parent education level; and (4) prior infant sibling number. The thresholds chosen for these stratification variables were: 1000 g, 21 years, college credits, and 1 prior infant sibling, respectively. An example of the forced-balance randomization is: when assigning a 20 year-old subject to a group, the algorithm searched all groups for subject age and, upon finding a group with more participants whose ages were less than 21 years, the likelihood of assigning the 20-year-old subject to that group was decreased.
Inclusion and exclusion criteria were used to ensure infant safety and parent follow-through. Inclusion criteria were (1) 1 parent of a 4-day-old infant aged 32 to 36 6-7 weeks postconception and (2) referral for physical therapy from a physician or nurse practitioner in the NICU or step-down nursery. Exclusion criteria were (1) parents unavailable for instruction, (2) parents already instructed on physical therapy positioning and handling techniques, (3) infants medically unstable for handling, (4) infants on a ventilator or nasal continuous positive airway pressure, (5) parents unable to participate in the intervention due to physical or emotional instability confirmed by a physician or social worker, (6) parents with blindness or physical limitations that would interfere with instruction or handling, or (7) parents not speaking or understanding English.
Subject recruitment and instruction occurred at Lucile Packard Children's Hospital Stanford in Palo Alto, California, or at one of the local satellite hospitals between September 2007 and October 2008. A convenience sample of parents whose infants met the criteria for referral to a physical therapist was used in the study. The study was designed to have an α of 0.05, with a power of 0.80. An a priori power analysis estimated that 84 participants were needed to achieve significant findings. Institutional Review Board approval was obtained for both the Rocky Mountain University of Health Professions and Lucile Packard Children's Hospital Stanford.
The handling competency measurement tool designed for this study was heavily influenced by Miller's pyramid of competency,12 consisting of 4 levels representing progressively increased aptitude. Miller's pyramid is both knowledge-based (“knows”) and performance-based (“shows”)13 and provides a 4-point Likert scale14 designed to assess ability in terms of “knows,” “knows how,” “shows how,” and “does.” These levels of proficiency are used to assess depths of learning, where “knows” represents the lowest level of learning and “does” the highest.15 Interventions in this study were focused on the parents’ ability to perform a simple whole motor task, and the assumption was made that if the parent can “show” or “do” accurately, then the parent “knows” and “knows how.” Therefore, the modified version of Miller's tool adapted for this study consisted of only the highest level of aptitude: “does.”
The modified tool used a 4-point Likert scale adapted from the scoring criteria by Perkins and colleagues16 to examine effective positioning in the NICU. This 4-point scale consisted of levels ranging from “no effort” to “ideal effort.” The scale was used in conjunction with a 10-point criterion checklist resulting in a low score of 10 and a high score of 40 for each of the 3 motor tasks (see Supplemental Digital Content 1, available at: http://links.lww.com/PPT/A242, Supplemental Digital Content 2, available at: http://links.lww.com/PPT/A243, and Supplemental Digital Content 3, available at: http://links.lww.com/PPT/A244).
Use of a visual analog scale (VAS) to measure the effectiveness of interventions in the medical setting is not novel.17 The VAS developed for this study consisted of 8 questions measured on a 6-point scale (1 = strong disagreement, 6 = strong agreement). The first 4 questions focused on the instruction itself; for example: was the instruction thorough and complete and was it clear and easy to understand? Questions 5 to 8 focused on the effect of the instruction (ie, “As a result of the instruction, I am more comfortable handling and positioning my baby”).
The Neonatal Medical Index (NMI)18 was completed for each infant to note the severity of infant illness prior to participation in the study. This index consists of 5 levels to indicate neonatal illness severity (1 = most benign, 5 = most complicated) associated with risk of neurodevelopmental complications (1 = low, 5 = very high). These data were used to analyze whether the severity of infant illness during the hospital course affected results.
Demographic information was collected to ensure balanced groups during randomization and for post hoc data analyses. The demographic data for this study included infant birth weight, parent age, parent education (primary school, some high school, some college, or some postcollege), number of prior infant siblings (any number of children reared by the parent since infancy), parent gender, and parent ethnicity.
Instructional Methods. Common instructional methods for physical procedural tasks include direct instruction, video demonstration, static presentation with pictures, verbal-only instruction, and written-only instruction. No one method of instruction can be deemed the best for teaching all procedural tasks because effective methods are likely task-environment and learner-dependent. The 3 methods chosen for this study were (1) direct instruction and demonstration by an investigator, (2) video with the same investigator performing the instruction and demonstration, and (3) written-pictorial instruction (see Supplemental Digital Content 4, available at: http://links.lww.com/PPT/A245) with pictures of the handling sequences. These instructional methods were chosen because they are feasible, commonly used in physical therapy, and supported in the literature. For standardization across groups, a written script was used to guide the language used during direct and video instructions and this same script was provided to the written-pictorial group.
Handling Tasks. The 3 handling activities used in this study were whole motor, procedural tasks. Procedural tasks apply sequenced actions to accomplish a whole task.19 The 3 handling tasks were e-mailed to 5 neonatal physical therapy experts from regions across the country for face validity. The feedback they provided was used when finalizing the instructional handouts. The activities included in this study were chosen because they are considered developmentally beneficial for infants born preterm.
The first activity was moving the infant into the side-lying positioning, which facilitates midline positioning of upper and lower extremities and chin tucking for the development of head control. Varying the position of the infant's head is important for maintaining musculoskeletal integrity20 (eg, preventing an acquired torticollis)21 as well as for cranial shaping.22 The side-lying facilitates the fine motor skills needed for hand to mouth and face play and hand to body play, activities that demonstrate motor-based self-regulatory behaviors.23 Placing neonates in the side-lying position contributes to coordinated spontaneous midline movements.24
The second handling activity was a guided lower extremity (LE) exercise performed with the infant in the supine position. This activity facilitated both free kicking and guided contained kicking. Infants who are born preterm demonstrate kicking movements in utero but are often limited by the constraints of body swaddling in the NICU. Flexion and extension movements of the legs may be beneficial to joint molding, and infants born preterm lose the fluid-filled environment presumably needed to promote the level of joint formation of infants born full term.20 By providing an opportunity for the infant to gently push against a parent's hand (similar to pushing against the intrauterine wall), these exercises may support bone mineralization.25
The third handling activity was a supported sitting task performed with the infant slightly flexed at the trunk, similar to a burping position. This activity may be beneficial because it facilitates the use of the neck rotator and extensor muscles, which fosters development of head control due to muscle strengthening.26 In addition, supported sitting promotes an alert and visually oriented infant for increased engagement with the caregiver.26
Rater Training and Reliability. The direct instruction and demonstration were conducted by the first author and 3 trained masked evaluators, who received 7 training sessions (2 with a doll; 1 with a newborn who was health and full term; and 4 with an infant who was healthy and preterm). During the 7 training sessions, each masked evaluator scored all items for each of the 3 handling activities and total scores and percent agreement were calculated. Intrarater agreement was 93% to 97% among the 3 evaluators. After the training sessions, the masked evaluators received written instructions for scoring parent competency and for conducting the follow-up session (ie, crib positioning, activity sequencing depending on initial infant positioning, approved parent prompts to keep session moving, criteria for stopping the session, and instructions for completing the session).
Parent Training and Testing. Baseline competency of the handling tasks was assessed by having all parents demonstrate the tasks using a preemie simulation doll (Children's Medical Ventures, Monroeville, Pennsylvania) when their infants were between 32 and 36 6/7 weeks postconception age in the NICU setting. To ensure consistency, the principal investigator followed a script for verbal prompting and rated each subject's preinstruction competency using a competency checklist designed for the study. After baseline competency assessment, 1 instructional session was provided in accordance with the assigned group: direct instruction, video, or written-pictorial. Instructional sessions were limited to 40 minutes and the following sequence was implemented: placement of the infant in a side-lying position; guided tactile and kinesthetic LE exercises; and facilitation of a supported-sit position. Parents were instructed to practice each activity with their infants 1 to 2 times per day for 2 to 4 minutes. All parents were provided the same written-pictorial handout to use for reference and were asked to maintain a log of their practice. Parents were instructed how to read infant cues and were advised to contain and soothe their infants during handling if needed. At the conclusion of the instructional session, parents rated their perceived effectiveness of the instruction on a VAS designed for the study, and a follow-up visit with the masked evaluator was scheduled for 2 to 4 days later.
During the follow-up visit with the masked evaluator, parents were asked to demonstrate each handling task using their infants. The postintervention testing occurred at the bedside. The masked evaluator ranked the parents’ performance of the handling activities using the same competency checklist used in the initial instructional session. Parents completed the VAS again to assess possible changes in their perceived effectiveness of instruction after independent practice. At the close of the session, the masked evaluator collected the practice log, answered questions, and offered parents an opportunity to receive an alternative form of instruction (no data collected).
To examine whether there was a change in performance measures, repeated-measures analyses of variance were performed to test the equality of means under numerous conditions. Using a general linear model, several repeated-measures analyses of variance were conducted with the pre- and posttest performance measures (supported sit competency, lower extremity [LE] kicking competency, and side-lying competency) as within-subject factors and instructional methods (direct, video, and written-pictorial) as between-subject factors. Using estimated marginal means for each performance measure, the NMI, infant birth weight, parent education, parent age, and number of prior infant siblings were examined as covariates for comparison between instructional methods. To examine differences between groups, t tests were calculated for the means on the VAS immediately following instruction and at the follow-up visit.
A total of 175 eligible participants were identified during the recruitment phase, of which 112 met criteria for participation, 96 agreed to participate and 86 participants completed the study. Demographic information is shown in Table 1. Instruction significantly altered task performance for all competency measures. The differences between pre- and postinstruction scores (Table 2) for each task were side-lying competency (t = 24.7, P = .01), guided LE competency (t = 16.3, P = .01), and supported sit competency (t = 14.1, P = .01) (Figure 1). Significant differences were found in postinstruction performance when instructional groups were compared. Direct and video instructions were significantly more effective than written-pictorial instruction (written vs direct side-lying competency P = .047, written vs video side-lying competency P = .034, written vs direct LE movement competency P = .007, written versus video for LE movement competency P = .018) for improving performance of the side-lying and guided LE activities (Figure 2). This trend was also observed for the supported sitting activity but failed to reach significance (written vs direct P = .083, written vs video P = .087). No significant differences were detected between the direct and video groups for the side-lying (P = .94) and guided LE (P = .89) tasks.
Immediately following the instruction, parents perceived the intervention positively (VAS scores from 5.4 to 5.8/6.0 on questions 1-4) and they expressed being more comfortable handling and positioning their infants (VAS scores from 5.1 to 5.7/6.0 on questions 5-8) (Table 3). No significant differences occurred between VAS scores after the initial instruction and the follow-up visit (direct P = .13, video P = .28, written P = .07).
Estimated marginal means for the NMI score, infant birth weight, parent education, parent age, and prior infant siblings were calculated to examine these variables as possible covariates.27 Data for covariates are included in Table 1. The mean NMI score was 2.4 ± 1.1. In this study, 50 infants had histories consistent with an NMI grade of 1 or 2, 32 infants had an NMI score of 3, and 14 infants had scores of 4 to 5. The mean birth weight of infants was 1612 ± 476 g and the mean parent age was 33 ± 6 years. The mean change of performance measures did not differ between instructional groups with different NMI score, infant birth weight, parent age, or number of prior infant siblings. The only demographic variable affecting results was parent education level, where 10% of participants had high school education or less, 45% of participants had some college, and the remaining 45% of participants had some postcollege.
Analyses of parent performance with the handling activities and parent education level as covariates revealed that parents with no college education (high school education or lower) scored lowest in their groups when written-pictorial or direct instruction was used. However, these parents scored highest in their group when video instruction was used. This was true for all 3 positioning and handling techniques. Conversely, those with a college education scored lowest on all 3 handling techniques when video instruction was used. In the written-pictorial instruction group, parents with a postcollege education scored the highest on all handling tasks. In addition, after instruction, parents with postcollege education scored significantly higher than parents with college education on the side-lying competency (F = 5.6, P = .05) and supported sitting competency (F = 4.45, P = .015).
Practice time did not significantly affect performance outcomes. The average numbers of times the activities practiced were 1.6 ± 1.6, 1.5 ± 1.5, and 1.7 ± 1.6 for the side-lying, guided LE movement, and support sit activities, respectively.
Facilitating optimal parent-infant interaction in a NICU is challenging, as many parents view the NICU as a foreign and intimidating environment. Fortunately, the physical therapist can help parents feel comfortable in interacting with and handling their infants. In this study, instruction in infant-handling and positioning techniques resulted in significant improvement in performance across 3 instructional methods and was positively perceived by parents. These results are not surprising and indicate that instructing parents to perform simple, whole motor tasks with their infants in the NICU is both affective and welcomed.
While performance across all instructional groups improved significantly, those in the written-pictorial group performed each of the 3 positioning and handling techniques less affectively than those in the direct and video instruction groups. It is therefore reasonable to conclude that the use of written instruction, even when presented with pictorial examples, is not an adequate teaching method when direct or video instruction is available. This may be because written-pictorial instruction lacks the dynamic and spatial relationships that may be important for learning a whole motor task. This has clinical importance because physical therapists are sometimes expected to leave instructional material at the bedside for patients or families without further instruction. The findings of this study suggest that this clinical practice may be insufficient for instruction of even simple, whole motor tasks in the NICU setting.
No significant differences in performance were demonstrated between the direct and video instruction groups. This result was not surprising because the handling and positioning activities in this study were simple, whole motor tasks, which could be sufficiently conveyed through a video demonstration. The minimal differences in performance between the direct and video instruction groups indicate that detailed video instruction may be a valuable tool for simple, whole motor tasks when direct instruction is unavailable.
Unlike the side-lying and LE activities, performance of the supported sit activity did not differ significantly between groups. In addition, the supported sit activity had the lowest postinstruction mean performance across all instructional groups. A possible reason for this unexpected finding may be that this activity was the most complex and, therefore, the most difficult in which to achieve competency.
Severity of the infants’ hospital courses did not influence results. Using the NMI, over half of the infants in the study had medical histories associated with a low to low-intermediate risk for serious neurobehavioral complications. The infants were 34 weeks and 4.6 ± 6.8 days postconception on average when studied. Handling techniques were taught in a context of fostering parent-infant interaction, and infants were developmentally mature enough for the handling activities. This combination of factors may explain this finding.
Infant birth weight was a stratification variable for randomization, due to concerns it could confound results. The result showing no effect on parent performance was unexpected and counters an assumption that parents may have more difficulty handling smaller infants. The number of infant siblings also did not significantly affect results and is commensurate with similar results reported by Satterfield and Yasumura.28 The finding that amount of practice did not significantly affect parent performance is likely due to the short time frame for follow-up.
Unlike the previously discussed covariates, the education level of participants did affect the results. Participants with more formal education performed best after receiving written-pictorial instruction while those with less formal education performed best after receiving video instruction. This finding is important and warrants further study.
Five methodological limitations were identified. The language of the participants was limited to English, and participant learning style was not assessed. In addition, long-term performance was not evaluated, and learning was measured only in the psychomotor domain. The last identified limitation was the lack of reliability and validity indicators in the measurement tools developed for the study.
Instructing parents to handle their infants in a developmentally sensitive manner in a NICU setting is effective for improving handling and positioning techniques. Direct and video instructions are equally effective for instructing parents how to perform simple whole motor tasks in the NICU setting, although parents with more formal education may learn better with written supplementation. Given limitations of time and personnel to provide direct instruction, members of the neonatal team should explore the utility of using more video instruction as a means of using personnel resources responsibly while increasing parent access to instruction.
The authors thank the masked evaluators Deborah Tong, PT, Heather McKersie, OT, and Nick St. John, PhD, members of the developmental team at Lucile Packard Children's Hospital Stanford. Sincere appreciation is expressed to Luba Botcheva, Stanford University statistician.
1. Reddick B, Catlin E, Jellink M. Crisis within crisis: recommendations for defining, preventing, and coping with stressors in the NICU. J Clin Ethics. 2001;12(3):254–265.
2. Shaw R, Bernard R, Deblois T, Ikuta L, Ginzburg K, Koopman C. The relationship between acute stress disorder and posttraumatic stress disorder in the neonatal intensive care
unit. Psychosomatics. 2009;50(2):131–137.
3. De Kloet ER. Hormones, brain, and stress. Endocr Regul. 2003;37(2):51–68.
4. Brown J, Talmi A. Family-based intervention to enhance infant-parent relationships in the neonatal intensive care
unit. J Pediatr Psychol. 2005;30(8):667–677.
5. Gale G, Franck L. Toward a standard of care for parents of infants in the neonatal intensive care
unit. Crit Care Nurse. 1998;18(5):62–74.
6. Beckwith L, Cohen SE, Kopp CB, Parmelee AH, Marcy TG. Caregiver-infant interaction and early cognitive development in preterm infants. Child Dev. 1976;47(3):579–587.
7. Sameroff A, Chandler M. Reproductive risk and the continuum of caretaking casualty. In: Horowitz F, Hetherington M, Scarr-Salapatek S, Siegel G, eds. Review of Child Development Research. Chicago, IL: University of Chicago Press; 1975.
8. McKim EM. The information and support needs of mothers of premature infants. J Pediatr Nurs. 1993;8:233–243.
9. Sweeney JK, Chandler LS. Neonatal physical therapy
: medical risks and professional education. Infants Young Child. 1990;2:59–68.
10. Sweeney JK, Heriza C, Blanchard Y, Dusing S. Neonatal physical therapy
. Part II: practice frameworks and evidence-based practice guidelines. Pediatr Phys Ther. 2010;22(1):2–16.
11. Efron B. Forcing a sequential experiment to be balanced. Biometrika. 1971;58(3):403–417.
12. Van Der Vleuten C. Validity of final examination in undergraduate medical training. BMJ. 2000;321(7270):1217–1219.
13. Schuwirth L, Van Der Vleuten C. The use of clinical simulations in assessment. Med Educ. 2003;37(suppl 1):65–71.
14. Norcini JJ. ABC of learning and teaching in medicine: work based assessment. BMJ. 2003;326(7392):753–755.
15. Rethans JJ, Norcini JJ, Baron-Maldonado M, et al The relationship between competence and performance: implications for assessing practice performance. Med Educ. 2002;36(10):901–909.
16. Perkins E, Ginn L, Fanning JK, Bartlett D. Effect of nursing education on positioning of infants in the neonatal intensive care
unit. Pediatr Phys Ther. 2004;16(1):2–12.
17. Edwards SGM, Playford ED, Hobart JC, Thompson AJ. Comparison of physician outcome measures and patients perception of benefits of inpatient neurorehabilitation. BMJ. 2002;324(7352):1493.
18. Korner AF, Stevenson DK, Kraemer HD. Prediction of the development of low birth weight preterm infants by a new Neonatal Medical Index. J Dev Behav Pediatr. 1993;14(2):106–111.
19. Leshin CB, Pollack J, Reigeluth CM. Instructional Design Strategies and Tactics. Englewood Cliffs, NJ: Educational Technology Publications; 1992.
20. Sweeney JK, Gutierrez T. Musculoskeletal implications of preterm infant positioning in the NICU. J Perinat Neonatal Nurs. 2002;16(1):58–70.
21. Neufeld S, Birkett S. What to do about flat heads: preventing and treating positional occipital flattening. Axone. 2000;22(2):29–31.
22. Najarian S. Infant cranial molding deformation and sleep position: implications for primary care. J Pediatr Health Care. 1999;13(4):173–177.
23. Grenier I, Bigsby R, Vergara E, Lester B. Comparison of motor self-regulatory and stress behaviors of preterm infants across body positions. Am J Occup Ther. 2003;57(3):289–297.
24. Nakano H, Konishi Y, Taga G, et al The effect of positioning on the general movements of preterm babies. Paper presented at: 14th Biennial International Conference on Infant Studies; May, 2004; Chicago, IL.
25. Litmanovitz I, Dolfin T, Friedland O, et al Early physical activity intervention prevents decrease of bone strength in very low birth weight infants. Pediatrics. 2003;112(1, pt 1):15–19.
26. Sheahan MS, Brockway NF. The high-risk infant. In: Tecklin JS, ed. Pediatric Physical Therapy
. 2nd ed. Philadelphia, PA: J.B. Lippincott Company; 1994.
27. Daniel W. Biostatistics: A Foundation for Analysis in the Health Sciences. 7th ed. New York, NY: John Wiley and Sons; 1999.
28. Satterfield MJ, Yasumura K. Facilitating the high risk neonate's head control: effect of teaching method on mother's performance. Int J Rehabil Res. 1987;10(1):55–62.
infant posture; neonatal intensive care; parent/education; physical therapy; teaching/methods
Supplemental Digital Content
Copyright © 2019 Academy of Pediatric Physical Therapy of the American Physical Therapy Association