Share this article on:

Responsiveness of the Test of Infant Motor Performance (TIMP) in Infants Born Preterm

Rose, Rachel Unanue PT, PhD, PCS; Westcott, Sarah L. PT, PhD

doi: 10.1097/01.PEP.0000176575.63915.67
Research Report

Purpose: The responsiveness of the Test of Infant Motor Performance (TIMP) was examined in infants born preterm. Infant variables predictive of changes in TIMP scores were also identified.

Methods: Twenty-five infants born ≤30 weeks’ gestational age were assigned a risk score and administered the TIMP at 32 and 36 weeks postconceptional age (PCA). Paired t-tests and analyses of variance examined differences between TIMP scores across time and between groups. Correlations and regression analyses determined the relationship between infant variables and TIMP score changes.

Results: TIMP scores significantly increased between 32 and 36 weeks PCA. A significant effect of time for risk groups was found. No significant correlations between infant variables and TIMP score changes were revealed.

Conclusions: The TIMP was responsive to changes in infants born preterm but did not discriminate based on risk classification. The infant variables did not predict changes in TIMP scores. Further research with a larger sample size is warranted.

The researchers found that the TIMP was responsive to changes in infants born preterm, but did not discriminate based on medical risk classification. They suggest continued study with a larger sample of infants.

The University of Alabama at Birmingham, Department of Physical Therapy (R.U.R.); and Programs in Rehabilitation Sciences, College of Nursing and Health Professions, Drexel University, Philadelphia, and Lake Washington School District, Redmond, WA, and Northwest Pediatric Therapies, Issaquah, WA (S.L.W.)

Address correspondence to: Rachel Unanue Rose, The University of Alabama at Birmingham, Department of Physical Therapy, RMSB 387, 1705 University Boulevard, Birmingham, AL 35294-1212. Email: UnanueR@uab.edu

Grant Support:This project was funded in part by the Maternal and Child Health Bureau Grant J-MCJ-429506-01 awarded to Drexel University for the preparation of pediatric physical therapists.

Back to Top | Article Outline

INTRODUCTION

The Test of Infant Motor Performance (TIMP) is a motor scale designed to assess both the motor and postural control of infants younger than four months of age. The TIMP can be used with both infants born preterm between 32 weeks postconceptional age (PCA) through four months corrected age and for infants born at term up to four months chronological age.1,2 The TIMP was developed to quantitatively assess motor development and identify infants whose development is outside the range of normal and might benefit from early intervention services.1,3,4

The TIMP, Version 3.3, is composed of 28 Observed Scale items and 31 Elicited Scale items.3 The Observed Scale items are scored based on observations of the infant’s spontaneous movements. The Elicited Scale items require handling to stimulate a motor or postural response. Although the TIMP has not been normed, data are available on typical performance on the TIMP for the different age ranges that can be used to identify infants who may benefit from follow-up or intervention.3

The TIMP has been tested to determine reliability and validity in infants up to four months of age.1,4,5 Test-retest reliability was demonstrated to be good across the age ranges of the TIMP (r = 0.89).5 In addition, the TIMP was examined for construct and concurrent validity.1,4 Based on the construct of postural control for functional mobility, scores on the TIMP discriminate infants of different ages and with different medical complications.1,2,6 TIMP items are also representative of the environmental demands experienced by infants born preterm and at term throughout their day.7 This finding of ecological relevance lends further support to the construct validity of the TIMP. In addition, at three months of age, the TIMP and Alberta Infant Motor Scales (AIMS) have similarities in their items and identify a similar group of infants with low motor performance supporting the concept of concurrent validity.4 The TIMP scores within the first three months also predict AIMS scores at six, nine, and 12 months, with the three-month TIMP scores showing the greatest predictive validity with the 12 month scores on the AIMS.8

If the goal of a measurement tool is to evaluate the effect of an intervention or to follow infants over time, evidence of the responsiveness of the measurement tool is needed. Responsiveness refers to the ability to measure clinically important change and is an essential property of an evaluative measure.9 A responsive measure not only detects changes in a specific population over time but also the ability to discriminate between two groups, one expected to make more changes and one expected to make fewer changes.10 Two longitudinal studies focusing on infants born preterm have shown the TIMP to be sensitive to the effects of a neurodevelopmental treatment program and a motor developmental program.11,12 The TIMP can also discriminate between a group of infants born preterm and infants born at term with different risks for motor delay.6 The TIMP has been found to be sensitive to the effect of intervention and can be used to discriminate between different groups of infants at risk for developmental delay.6,11,12 However, the evidence does not specifically address whether the TIMP is responsive to changes in infants born preterm or can be used to discriminate between two groups of infants born preterm rated at higher and lower risk for developmental delays.

The assessment of postural control in infants born preterm is important because it may be the rate-limiting factor in the acquisition of motor milestones.13,14 Assessing development of postural control may help identify infants in need of services. Another important aspect of assessing infants born prematurely is being able to identify impairments associated with the changes in motor and postural control in infants born preterm. In addition, identification of medical markers or impairments that may predict scores on the TIMP at later follow-up is also important for early identification of infants at risk for developmental delays. The authors of one study has found correlations between the scarf ratio, an assessment of upper extremity muscle tone, with TIMP scores at 40 weeks PCA and one, two, three, and four months adjusted age.15 In addition, the scarf ratio was also correlated to specific TIMP items related to posture and upper extremity movement. Although the scarf ratio may be a good indicator of later motor performance on the TIMP, no evidence exists to determine whether specific medical complications or infant characteristics such as birth weight, gestational age, Apgar scores, presence of an intraventricular hemorrhage (IVH) or periventricular leukomalacia (PVL), seizures, and referral to physical and/or occupational therapy while in the Neonatal Intensive Care Unit (NICU) predict changes on the TIMP. These variables are important due to their influence on an infant’s ability to show developmental changes over time.16–21

The purposes of this study were to examine the TIMP’s responsiveness to change in infants born preterm and to identify variables predictive of changes in TIMP scores. The hypotheses for this study were as follows: (1) The TIMP will be responsive to motor and postural control change in infants born preterm and will discriminate between infants at higher and lower medical risk and (2) infant variables including gestational age, birth weight, Apgar scores, IVH, PVL, seizures, and referral for therapy will predict changes in TIMP scores. It was hypothesized that infants with at least one of the following characteristics will show less change on the TIMP over time than infants who do not have these characteristics: lower gestational age, low birth weight, low Apgar scores, IVH, PVL, seizures, and those who receive therapy.

Back to Top | Article Outline

METHODS

Subjects

A sample of convenience of 32 infants was recruited from the NICU at St. Louis Children’s Hospital and the Special Care Nursery (SCN) at Barnes-Jewish Hospital in St. Louis. Infants were recruited as part of a larger study involving examination of a parent education program. Infants born at or before 30 weeks’ gestational age with one parent who visited the NICU or SCN at least three times per week were eligible for inclusion in this study. Infants were excluded if they had a genetic disorder, known congenital anomalies, cardiac defects other than a patent ductus arteriosus, prolonged sedation or muscle paralysis (defined as greater than or equal to one week), maternal history of substance abuse, and parents younger than the age of 18 years. Informed consent approved by the Human Subjects Committee at Washington University and the Internal Review Board at MCP Hahnemann University was obtained for all infants prior to their entrance into the study.

Twenty-five infants, 14 female and 11 male infants, completed the study. Seven infants were withdrawn due to transfers to outside hospitals, early discharge, or the parent’s failure to meet with the physical therapist providing parent education. The charts of all infants were reviewed for specific medical history for descriptive purposes and to determine a risk score for each infant. Six infants were born at or before 26 weeks’ gestational age and 19 were born between 27 and 30 weeks’ gestational age. Infants’ birth weights ranged from 420 to 1524 g. Nineteen infants had oxygen requirements when first assessed. See Table 1 for demographic and risk variables.

TABLE 1

TABLE 1

Back to Top | Article Outline

Risk Classification:

All infants were classified as at higher or lower risk for neurodevelopmental delays based on the Neurobiological Risk Score (NBRS).22 The NBRS has been shown to be a reliable risk score focused on the mechanisms of brain cell injury.22,23 The NBRS has also been shown to be a good predictor of developmental outcome throughout the first 24 months of development.22,23 The NBRS determines infant risk through information from the infant’s medical chart. Seven items comprise the NBRS: ventilation, pH, seizures, IVH, PVL, infection, and hypoglycemia. Item scores are weighted for severity and duration, taking into account cumulative events. Items are scored 0 if the risk item is absent and 1, 2, or 4 if the item is present.22–24 Total scores of 8 or above designate high risk, scores between 5 and 7 are intermediate risk, and total scores at 4 or below designate low risk.23 For the purposes of this study, all infants were classified using the NBRS as either lower risk (scores ≤4) or higher risk (scores ≥5). The infants who fell into the intermediate or high risk group as defined by Brazy et al23 were grouped together as higher risk due to the small numbers of infants in those classifications. Previous research has shown infants scoring a 4 on the NBRS were not at any greater risk of neurodevelopmental delays than those scoring at 3 or below. However, infants scoring a 5 had a much higher incidence of neurodevelopmental delays.23 Based on this information, the intermediate and high risk groups, infants scoring above 5, were collapsed into one group of infants at greater risk of neurodevelopmental delays. The infants who participated in this study had risk scores ranging from 1 to 8. Nineteen infants were classified as lower risk and six as higher risk.

Back to Top | Article Outline

Raters

One physical therapist and two occupational therapists with pediatric and NICU experience were trained to administer the TIMP. All three trainees reviewed a self-study CD-ROM training program, practiced administering and scoring the TIMP, and established interrater reliability set by TIMP authors using criterion videos. Once reliability was established, each trainee was videotaped performing the TIMP on two infants and received written feedback on their administration of the TIMP from a research assistant involved in the development of the TIMP CD-ROM training program. The interrater reliability between the three testers was checked throughout the study using percentage of agreement due to the small number of subjects. Item reliability ranged from 67% to 100%. Three of the 31 items on the Elicited Scale had 67% agreement between raters. For all other items, reliability was 83% to 100%.

Back to Top | Article Outline

Procedure

The TIMP was administered to all 25 infants at 32 weeks PCA and then again approximately four weeks later at 36 weeks PCA or discharge, whichever occurred first. Each infant was assessed by one of the three raters trained in the administration of the TIMP. Throughout each testing session, heart rate, oxygen saturation, and behavioral cues were observed and recorded for any signs of stress in the infants. All testing was completed in one session with rests as needed due to the length of the test session. There were no increases in heart rate, heart rate drops, or decreases in oxygen saturation below recommended levels during any of the test sessions. Recommended levels were heart rates between 100 and 200 beats per minute and oxygen saturation greater than 90%.

Back to Top | Article Outline

Data Analysis

A Pearson product moment correlation was calculated to determine whether TIMP scores at 32 weeks PCA were associated with TIMP scores at 36 weeks PCA. A paired t test was performed on TIMP scores at 32 weeks PCA and 36 weeks PCA to determine whether there was a significant difference in mean scores. A group-by-time repeated-measures analysis of variance (ANOVA) was conducted to determine whether there was a difference between the higher and lower risk classification groups on the TIMP.

Pearson product moment correlations were conducted to determine whether infant variables were correlated with one another and with TIMP change scores. A hierarchical regression analysis was performed between the change scores on the TIMP across the four weeks and those infant variables correlated with the TIMP change scores. If two infant variables were highly correlated with one another, only one was considered for entry into the hierarchical regression. An alpha level of 0.05 was used to determine statistical significance for all analyses.

Back to Top | Article Outline

RESULTS

The correlation between TIMP scores at 32 and 36 weeks PCA was not significant (r = −0.287, p = 0.165). The distributions of the TIMP scores at 32 and 36 weeks PCA are presented in Figure 1. The infants’ mean TIMP score at 32 weeks PCA was 26.9 (SD = 8.7) and their mean TIMP score at 36 weeks PCA was 68.2 (SD = 16.7), an improvement that was statistically significant (t = −9.9, p = 0.0001).

Fig. 1.

Fig. 1.

The distributions of the TIMP scores at 32 weeks PCA and 36 weeks PCA based on risk classification are presented in Figure 2. The mean TIMP score at 32 weeks PCA for the lower risk group was 26.3 (SD = 8.9) and for the higher risk group was 28.8 (SD = 8.2). The mean TIMP score at 36 weeks PCA was 70.4 (SD = 17.9) for the lower risk group and 61.2 (SD = 10.5) for the higher risk group. The results of the repeated-measures ANOVA (Table 2) indicate a significant effect of time, but no significant group effect or interaction of group and time. The results indicate that both risk classification groups demonstrated improvements in TIMP scores between 32 and 36 weeks PCA, but there were no significant differences between the two groups.

Fig. 2.

Fig. 2.

TABLE 2

TABLE 2

The results of the correlations with the TIMP change scores are listed in Table 3. The correlations between each infant variable and TIMP change scores were low, ranging from 0.02 and 0.38, and were not statistically significant. Only therapy services which had the highest correlation, r = 0.38, were entered into a regression analysis. Fourteen percent of the variance in TIMP change scores was accounted for by therapy services. The results of the regression analysis are shown in Table 4.

TABLE 3

TABLE 3

TABLE 4

TABLE 4

Back to Top | Article Outline

DISCUSSION

Responsiveness measures change over time and as does the ability to differentiate between high- and low-risk groups. The total group of infants significantly increased their mean TIMP scores between 32 and 36 weeks PCA. In addition, the smaller groups of higher and lower risk infants significantly increased their TIMP scores across the four weeks. However, the TIMP did not discriminate between lower and higher risk groups, those who would potentially make greater change and those who would potentially make less change.

The results of this study are in contrast to those of Campbell and Hedeker6 in which the TIMP discriminated infants based on risk for motor delays. Using the infant classification of Campbell and Hedecker,6 all 25 infants participating in this current study would be classified as high risk based on gestational age or birth weight. All infants in this study were infants born preterm and not a mix of infants that included low-risk infants born at term and medium- and high-risk infants born preterm. The current study appears to be comparing the two ends of a high-risk group, which may account for why there were no differences between the two risk classification groups.

The findings that the change in TIMP score did not differ between infants at lower medical risk and infants at higher medical risk may have been affected by the small sample size and uneven risk group distributions. The lower risk group was three times larger than the higher risk group. A more even distribution may have shown a difference in TIMP scores based on the risk classification scores. In this study, only six infants were classified as higher risk. Of these six infants, five of the infants were actually classified as intermediate risk based on the NBRS classification by Brazy and colleagues.23 Because the two groups, intermediate and high risk, were collapsed into one higher risk group, the differences between the higher and lower risk groups may have been reduced. Infants who are classified as high risk based on the NBRS classification system may show significant differences on the TIMP compared to the infants classified as low risk. A larger sample size with subjects evenly distributed across the three classification groups is recommended to better examine the ability of the TIMP to discriminate infants based on the NBRS classification.

The risk classification used in this study does not include impairments that are directly linked with an infant’s changes in motor and postural control over time. The NBRS is based on medical risk factors documented in the chart. Impairments in muscle tone, motor patterns, and behavioral regulation may affect an infant’s ability to show change on the TIMP because these impairments can affect movement.25–30 In this study, some of the infants at lower risk had impairments in muscle tone and movement patterns that may have limited changes in postural and motor control. As noted above, only six infants were classified as higher risk; however, 16 infants were recommended for and received therapy while in the NICU or SCN. Two additional infants were referred to physical and occupational therapy at the end of this study. The 16 infants referred for therapy had been identified behaviorally as being at higher risk for developmental delays based on either nurses’ or physicians’ concerns about impairments they were seeing. These infants may already have shown increases or decreases in muscle tone, decreased active movements, decreased levels of alertness, or decreased tolerance to handling and stimulation. Any or all of these impairments may influence the infant’s abilities. The NBRS is appropriate for medically classifying infants at risk for delays; however, there may be a stronger relationship between infants who have been identified to have specific motor problems and the change in TIMP scores.

Results of this study did not support the hypothesis that the infant variables associated with motor delays would predict change on the TIMP. None of the infant variables chosen for their effect on development were correlated significantly with the change in TIMP scores. The small sample size could have affected the correlations between the different infant variables and the change scores on the TIMP. Although the correlation between the TIMP change scores and therapy services was not statistically significant, therapy services accounted for 14% of the TIMP change scores. Clinically, 14% is a significant amount of variance for therapy services in a small group. With a larger sample size, 14% of the variance may have been statistically significant in addition to clinically significant.

Two previous studies support the use of the TIMP to discriminate between those who received a treatment program and those who did not. Infants born preterm with motor delays who received neurodevelopmental treatment made a greater change on the TIMP compared to the control group of infants born preterm who did not receive intervention.11 In addition, Lekskulchai and Cole12 provide evidence of the sensitivity of the TIMP to the effects of a motor developmental program. Infants at risk for developmental delay receiving a motor developmental intervention showed greater improvements in their TIMP scores compared to the infants in the control group.

Measures of impairments may be significantly correlated with changes in TIMP scores, and therefore, improve the ability to predict change scores on the TIMP. These impairments may include measures of tone (scarf sign, arm recoil, or popliteal angle), measures of motor patterns, and signs of behavioral regulation. Impairments rather than medical risk factors may have a stronger relationship to changes that infants make on the TIMP. Further research is necessary to explore this issue.

Back to Top | Article Outline

CONCLUSIONS

The results provide evidence that infants born preterm demonstrate significant changes on the TIMP between 32 and 36 weeks PCA. Therefore, the TIMP is an appropriate measurement tool for use with infants born preterm as early as 32 weeks PCA to assess changes in motor and postural control. Change on the TIMP did not discriminate between infants classified as higher risk and infants classified as lower risk for future developmental delay. However, due to the small number of infants in this study and the method to designate higher and lower risk groups, further research is warranted. The selected variables associated with development did not predict change on the TIMP. Referral to physical therapy while in the NICU did show the highest correlation with the change in TIMP scores. Further research using a larger sample size is necessary to examine other variables that might be related to change in TIMP scores such as muscle tone and motor patterns.

Back to Top | Article Outline

ACKNOWLEDGMENTS

We thank Rebecca Lipnick and Catherine Genetti for their role in testing infants. We also thank the physicians and nurses in the NICU at St. Louis Children’s Hospital and the SCN at Barnes-Jewish Hospital in St. Louis for providing access to their patients. A special thank you to all the infants and their parents who participated in this study.

Back to Top | Article Outline

REFERENCES

1. Campbell SK, Kolobe THA, Osten ET, et al. Construct validity of the Test of Infant Motor Performance. Phys Ther. 1995;75:585–596.
2. Campbell SK, Osten ET, Kolobe THA, et al. Development of the Test of Infant Motor Performance. Phys Med Rehabil Clin N Am. 1993;4:541–550.
3. Campbell SK. The infant at risk for developmental disability. In: Campbell SK, ed. Decision Making in Pediatric Neurologic Physical Therapy. New York: Churchill Livingstone;1999:260–332.
4. Campbell SK, Kolobe THA. Concurrent validity of the Test of Infant Motor Performance with the Alberta Infant Motor Scale. Pediatr Phys Ther. 2000;12:2–9.
5. Campbell SK. Test-retest reliability of the Test of Infant Motor Performance. Pediatr Phys Ther. 1999;11:60–66.
6. Campbell SK, Hedeker D. Validity of the Test of Infant Motor Performance for discriminating among infants with varying risk for poor motor outcome. J Pediatr. 2001;139:546–551.
7. Murney ME, Campbell SK. The ecological relevance of the Test of Infant Motor Performance elicited scale items. Phys Ther. 1998;78:479–489.
8. Campbell SK, Kolobe THA, Wright BD, et al. Validity of the Test of Infant Motor Performance for prediction of 6-, 9-, and 12-month scores on the Alberta Infant Motor Scale. Dev Med Child Neurol. 2002;44:263–272.
9. Guyatt G, Walter S, Norman G. Measuring change over time: assessing the usefulness of evaluative instruments. J Chron Dis. 1987;40:171–178.
10. Stratford PW, Binkley JM, and Riddle DL. Health status measures: strategies and analytic methods for assessing change scores. Phys Ther. 1996;76:1109–1123.
11. Girolami GL, Campbell SK. Efficacy of a neuro-developmental treatment program to improve motor control in infants born prematurely. Pediatr Phys Ther. 1994;6:175–184.
12. Lekskulchai R, Cole J. Effect of a developmental program on motor performance in infants born preterm. Aust J Physiother. 2001;47:169–176.
13. Case-Smith J. Postural and fine motor control in preterm infants in the first six months. Phys Occup Ther Pediatr. 1993;13:1–17.
14. Shumway-Cook A, Woollacott M. Theoretical issues in assessing postural control. In: Wilhelm IJ, ed. Physical Therapy Assessment in Early Infancy. New York: Churchill Livingstone; 1993:161–171.
15. Lekskulchai R, Cole J. The relationship between the scarf ratio and subsequent motor performance in infants born preterm. Pediatr Phys Ther. 2000;12:150–157.
16. Collin MF, Halsey CL, and Anderson CL. Emerging developmental sequelae in the ‘normal’ extremely low birth weight infant. Pediatrics. 1991;88:115–120.
17. Nelson KB, Ellenberg JH. Neonatal signs as predictors of cerebral palsy. Pediatrics. 1979;64:225–232.
18. Nelson KB, Ellenberg JH. Apgar scores as predictors of chronic neurologic disability. Pediatrics. 1981;68:36–44.
19. Patel J, Edwards AD. Prediction of outcome after perinatal asphyxia. Curr Opin Pediatr. 1997;9:128–132.
20. Sinha SK, D’Souza SW, Rivlin E, et al. Ischaemic brain lesions diagnosed at birth in preterm infants: clinical events and developmental outcome. Arch Dis Child. 1990;65:1017–1020.
21. Volpe JJ. Brain injury in the premature infant: neuropathology, clinical aspects, pathogenesis, and prevention. Clin Perinatol. 1997;24:567–587.
22. Brazy JE, Eckerman CO, Oehler JM, et al. Nursery neurobiological risk score: important factors in predicting outcome in very low birth weight infants. J Pediatr. 1991;118:783–792.
23. Brazy JE, Goldstein RF, Oehler JM, et al. Nursery neurobiologic risk score: levels of risk and relationships with nonmedical factors. J Dev Behav Pediatr. 1993;14:375–380.
24. Oehler JM, Goldstein RF, Catlett A, et al. How to target infants at highest risk for developmental delay. Am J Matern Child Nurs. 1993;18:20–23.
25. Georgieff MK, Bernbaum JC. Abnormal shoulder girdle muscle tone in premature infants during their first 18 months of life. Pediatrics. 1986;77:664–669.
26. Georgieff MK, Bernbaum JC, Hoffman-Williamson M, et al. Abnormal truncal muscle tone as a useful early marker for developmental delay in low birth weight infants. Pediatrics. 1986;77:659–663.
27. Harris SR. Early neuromotor predictors of cerebral palsy in low-birthweight infants. Dev Med Child Neurol. 1987;29:508–519.
28. Harris SR. Movement analysis—an aid to early diagnosis of cerebral palsy. Phys Ther. 1991;71:215–221.
29. Prechtl HFR, Einspieler C, Cioni G, et al. An early marker for neurological deficits after perinatal brain lesions. Lancet. 1997;349:1361–1363.
30. Touwen BCL. Variability and stereotypy of spontaneous motility as a predictor of neurological development of preterm infants. Dev Med Child Neurol. 1990;32:501–508.
Keywords:

infant/premature; developmental disabilities/diagnosis; motor skills disorders/diagnosis; physical therapy/specialty/instrumentation; longitudinal studies

© 2005 Lippincott Williams & Wilkins, Inc.