Secondary Logo

Journal Logo


Changes in Therapist Actions During a Novel Pediatric Physical Therapy Program: Successes and Challenges

Akhbari Ziegler, Schirin PT, MSc; Dirks, Tineke PT; Reinders-Messelink, Heleen A. PhD, MA; Meichtry, André PT, MSc; Hadders-Algra, Mijna MD, PhD

Author Information
doi: 10.1097/PEP.0000000000000509


COPing with and CAring for Infants With Special Needs (COPCA) is a novel family-centered early intervention program.1–5 Theoretical background and practical performance of COPCA differ from approaches like neurodevelopmental treatment (NDT).1 COPCA is primarily based on principles and elements of family-centered care1,4,6,7 and the neuronal group selection theory (NGST).8–10 Goal-oriented coaching11 of the family is COPCA's main intervention strategy.1,4 COPCA has 2 objectives: (1) to empower the individual family in processes of decision-making regarding participation in daily life, and (2) to optimize motor capacities of the infant with special needs. Family-centered practice in COPCA is characterized as follows: (1) each family is regarded as unique, autonomous, and responsible; the family's values and choices are respected, and its strengths and needs considered; (2) the relationship between family and COPCA coach is based on equal partnership; (3) the family is constant in the child's life and caregivers are the experts on the child's needs.1,12,13 According to the NGST, typical motor development is characterized by variation and development of adaptability, the latter being based on the ability to select the most effective solution for a given task from the available repertoire. The selection is based on self-produced motor behavior and trial-and-error experiences.8–10 Atypical motor development is characterized by limited variation and limited adaptability. The latter is induced by impaired sensory processing and limited repertoire. Therefore, the infant with atypical motor development needs ample opportunities for variation of self-produced motor behavior and trial-and-error experiences to improve adaptability.9 Practical implementation of the NGST means to challenge the infant at the limit of the infant's capabilities to actively explore varying situations, which offer the infant opportunities for self-produced motor behavior and trial-and-error experiences. Goal-oriented coaching includes support of the family by shared observations, observation of the interplay between caregiver and infant during daily care, listening, asking specific questions, information exchange, and provision of guidance to challenge the infant. During shared observation, family members may explore and develop strategies to cope in daily life with having an infant with special needs. The caregivers are informed that development proceeds by means of trial-and-error and self-produced activity requiring ample time and practice. They also are informed that error does not mean failure but expresses the presence of an active learning process. In COPCA, the roles of the caregivers are as actively involved goal setters, decision–makers, and supporters of the child. During the intervention they are engaged in naturally occurring parenting situations (eg, playing or bathing). The therapist takes on the role of a coach, acts as an observer and listener, receives and transmits information, and facilitates the ideas and actions of family members. The first study on the effectiveness of COPCA in infants at high risk of developmental disorders supported that element of the COPCA approach, in particular caregiver coaching and challenging the infant to independently perform motor actions, was associated with a better functional outcome at 18 months.2,3

Implementation of new approaches bears similarities to implementation of evidence-based practice, as both depend on behavior changes of professionals,14 in this context pediatric physical therapists. The reason that evidence-based guidelines are often not implemented effectively may be that theories of human behavior are insufficiently taken into account.15 One way to improve implementation of a new approach is to explore the ways in which the approach is implemented16 (ie, to become aware of required behavior changes). For the implementation of COPCA, changes in the following domains, as described by the Theoretical Domains Framework,15,17 are needed: “changes in knowledge” (knowing underlying theories), “skills” (listening, asking specific questions, and providing of guidance), “social/professional role and identity” (role of coach), “goals” (expanded to the whole family), “environment” (at the family's home), “intentions” (stability of intentions), and “behavior regulation” (action planning, breaking habits).17(p13-14) The first study comparing the effectiveness of COPCA and other approaches2 supported that, in one third of the infants in the COPCA group, the COPCA principles were applied incompletely. This suggests that the COPCA coaches failed to accomplish all required behavior changes.

In Switzerland, the COPCA program is largely unknown. Pediatric physical therapists (PPTs) are typically trained in infant physical therapy (TIP) (eg, NDT) and spend most of treatment time using hands-on techniques and caregiver training. As we planned a study on the effectiveness of COPCA in Switzerland, a group of Swiss PPTs participated in a COPCA course. This offered an opportunity to monitor the implementation of COPCA in Switzerland. We hypothesized that behavior changes in the domain of “behavior regulation” are most difficult to achieve, as habits involve limited cognitive attention and therefore are less easily affected by a conscious change. The aim of this study is to describe the changes in time spent on pediatric physical therapy (PPT) actions and situations during the COPCA course by comparing the contents of PPT sessions just prior to, during, and shortly after completion of the course.


The data were collected just prior to, during, and after completion of a COPCA course, during which 15 PPTs received education to become a COPCA coach. The course was held in Groningen, the Netherlands, and coordinated by Mijna Hadders-Algra and Tineke Dirks, the developers of the COPCA program. Heleen Reinders-Messelink, a human movement scientist familiar with the COPCA program through participation in a recent randomized controlled trial on COPCA,5 was an additional teacher. The course started in May 2015 with 2 days of contact teaching (part I) followed by 1 day of contact teaching in September (part II) and 1 in November 2015 (part III). Contents of the contact days were lectures on theories underlying COPCA (eg, family-centered practice, the NGST, and goal-oriented coaching), basic principles of COPCA, and practical implementation through coaching and discussion on the basis of videos of treatment sessions performed by the participants. Each participant received course feedback on a video of one of her recent PPT sessions during part II and part III. These feedback sessions were 20 minutes per video. Feedback was provided both by teachers and peers. In addition, there were 2 supervision sessions of 30 minutes each during each interval between the course parts. A recent video-recorded PPT session of the participant was discussed. Supervision was provided remotely to teachers via Skype. Supervision focused on identifying the goals for the intervention process with the family (eg, applying the appropriate neuromotor action to reach the goal), family-centered practice (eg, participation of the autonomous, responsible, and individual family in the intervention process), application of the NGST (eg, to challenge self-produced motor behavior and trial-and-error experiences), and application of goal-oriented coaching (eg, asking specific questions, information exchange, and provision of hints). Teachers prompted participants to reflect on the intervention sessions by asking specific questions, and they provided guidance about the application of the COPCA principles.

Between April and December 2015, 1 child of each therapist received 1 treatment session per week. Prior to the COPCA course, treatment was in an outpatient clinic or a specific PPT setting. After part I, the intervention took place at the child's home. Six treatment sessions of each child were video recorded. The video recordings served as an instrument for supervision of the participants. The first session occurred just prior to the first part of the COPCA course, 2 sessions were recorded between the first and the second parts, and another 2 between the second and third parts of the course. Finally, the last video recording was made 2 weeks after the course completion.

Recruitment and Informed Consent

The PPTs were recruited in the German-speaking area of Switzerland in 2014 through phone or e-mail. After agreeing to participate, they received further written information regarding the study. In March and April 2015, the participating PPTs informed eligible families about the course and its associated study and selected 1 family with a young child with special needs. The caregivers and the PPTs signed a letter of consent. The ethics committee of the canton Zurich in Switzerland assessed the study protocol and decided that no further ethical approval was needed according to human research law, as the aim of the study was to evaluate the changes in the behaviors of the PPTs and not the effect of intervention on child development. However, the ethics committee recommended obtaining informed consent, as we did.


Pediatric Physical Therapists. The inclusion criteria for the PPTs were readiness to complete the COPCA course in the Netherlands and willingness to participate as COPCA coaches in a randomized controlled trial, evaluating the effect of COPCA and TIP on family outcome and infant development for infants at risk for developmental disorders. Fifteen female PPTs participated, which corresponded to the anticipated need in the randomized controlled trial. Background information on the PPTs is shown in Table 1.

TABLE 1 - Characteristics of Therapists and Children
Characteristics of therapists (n = 15)
Male/female, n 0/15
Country of basic PT education, n (%)
Switzerland 9 (60.0)
Germany 3 (20.0)
The Netherlands 2 (13.3)
Austria 1 (6.7)
Highest level of education, n (%)
Vocational school 3 (20.0)
Bachelor of sciences 8 (53.3)
Master of sciences or advanced studies 4 (26.7)
Overall professional experience, y
Median (range) 23 (1.5-38)
<1, n (%) 0 (0.0)
1-5, n (%) 1 (6.7)
5.1-10, n (%) 0 (0.0)
10.1-15, n (%) 2 (13.3)
>15, n (%) 12 (80.0)
Professional experience in infant PPT, y
Median (range) 15 (0-36)
<1, n (%) 1 (6.7)
1-5, n (%) 2 (13.3)
5.1-10, n (%) 1 (6.7)
10.1-15, n (%) 4 (26.7)
>15, n (%) 7 (46.7)
Characteristics of the infants/children (n = 15)
Male/female, n 7/8
Gestational age, wk
Median 39
Range 25-40
Preterm, n (%) 5 (33.3)
Birth weight, g
Median 3080
Range 790-4100
Age at first video recording
≤1y, n (%) 8 (53.3)
>1y, n (%) 7 (46.7)
Median, mo 9
Range, mo 2-35
Severity of disorder
GMFCS I or mild-to-moderate disorder, n (%) 9 (60)
GMFCS IV or severe disorder, n (%) 6 (40)
Abbreviations: GMFCS, Gross Motor Function Classification System; PPT, pediatric physical therapy; PT, physical therapy.

Children. Fifteen children 2 to 35 months' corrected age (median 9 months) with delayed motor development and/or neurological dysfunction and not able to stand and walk independently were included in the study. Background information on the children is shown in Table 1.

Video Analyses

Four of the 6 video-recorded treatment sessions were analyzed quantitatively with the Groningen Observation Protocol version 2 (GOP 2.0)18,19 and the software program, The Observer XT (version 11.5; Noldus, Wageningen, the Netherlands). The first video was from the session just prior to part I of the COPCA course (T0), the second and third videos were from the second sessions in the first and second intervals, respectively (T1 + T2), and the fourth was from the session recorded after the course completion (T3). The 60 recordings were randomized and masked to the chronology of recording time.

The GOP 2.0 defines therapeutic behaviors in 5 main categories: “educational actions toward caregiver,” “communication” (eg, “instruct” and “provide feedback”), “neuromotor actions” (eg, “facilitation techniques,” “challenge to self-produced motor behavior (CSPMB),” “position” of the infant, and “situation” (eg, “motor activity/play” and “dressing”). The main categories are further specified by subcategories, the behaviors (eg, “communications” are specified by “instruct,” “provide feedback” and “neuromotor actions” by “facilitation techniques,” and “CSPMB”). These behaviors are further specified by modifiers (eg, “instruct” may be further specified by the modifiers “gives strict instruction” or “gives hints”)1,18,19 (see Supplemental Digital Content 1, available at:

The GOP 2.0 quantifies the relative duration of therapeutic actions (percentage of total treatment time) during a session.19 The GOP 2.0 also allows for the recording of parameters unrelated to time, such as family members involved or the role of the caregiver. The GOP 2.0 is a reliable tool to assess the content of infant PPT in Switzerland.18 Interrater reliability between 2 raters was determined with a random sample of 5 videos with interclass coefficients (2) of 0.48 to 1.00 (median value 0.92).18 According to Shrout and Fleiss,20 values more than 0.75 reflect excellent agreement, those between 0.4 and 0.75 sufficient to good agreement, and all values less than 0.4 insufficient agreement.

Data Analyses

Characteristics of PPTs and children were quantified by descriptive statistics. Time spent with specified PPT actions and situations was quantified with medians and ranges of relative duration of treatment sessions. The Wilcox signed rank test was used to analyze changes in PPT actions over time. Median changes and corresponding nonparametric confidence intervals were computed using the Hodges-Lehmann estimator.21 All statistical analyses were done using SPSS (version 24). To prevent a type I error, P < .01 (99% confidence interval) was the cutoff for statistical significance.


Characteristics of Participating Pediatric Physical Therapists and Children

The PPTs had 15 years of professional experience in treating infants (median value; range 0-36 years). Nine of the PPTs obtained basic physical therapy education in Switzerland, 3 in Germany, 2 in the Netherlands, and 1 in Austria. Thirteen PPTs were trained in infant treatment according to NDT. All PPTs attended the 4 contact days, and individual supervision sessions; all completed the home assignments.

Median age of the children at the first video recording was 9 months' corrected age (range 2-35 months). Nine children had been referred to therapy because of delayed motor development with hypotonia, 4 because of cerebral palsy, and 1 because of muscular dystrophy and 1 because of epilepsy with delayed motor development.

Changes in Time Spent on PPT Actions

The treatment sessions at T0 were in private physical therapy or outpatient clinics. From T1 onward, sessions were at the infant's home. The duration of treatment sessions varied from 20 to 54 minutes (median value 35 minutes); duration did not change during the course. Table 2 includes time spent with actions and situations at T0, T1, T2, and T3.

TABLE 2 - PPT Situations and Actions Over Timea
PTT Situation or Actions T0 Median (Range) T1 Median (Range) Median Changes T0-T1 HL Estimate (99% CI) T2 Median (Range) Median Changes T0-T2 HL Estimate (99% CI) Median Changes T1-T2 HL Estimate (99% CI) T3 Median (Range) Median Changes T0-T3 HL Estimate (99% CI) Median Changes T2-T3 HL Estimate (99% CI)
Educational actions toward caregiver
Caregiver training 42.5 (0.0-100.0) 0.0 (0.0-56.5) 40.2 (−73.2 to 0.0) 0.0 (0.0-1.4) 45.8 (−81.7 to 0.0) 0.0 (−0.5 to 0.0) 0.0 (0.0-0.0) 45.8 (−81.7 to 0.0) 0.0 (0.0 to 0.0)
Caregiver coaching 0.0 (0.0-0.0) 93.3 (0.0-100.0) 77.9 (48.7 to 99.1) 98.5 (14.1-10.0) 92.7 (57.1 to 100.0) 10.2 (−20.2 to 46.2) 98.8 (3.5-100.0) 96.1 (51.7 to 100.0) 0.0 (−17.3 to 33.5)
Not specified educational actions 57.5 (0.0-100.0) 2.6 (0.0-100.0) 33.5 (−66.3 to −1.2) 1.5 (0.0-85.9) 41.9 (−78.7 to −8.3) −3.4 (−43.1 to 24.2) 1.2 (0.0-96.5) 46.7 (−87.8 to −3.3) 0.0 (−33.5 to 17.3)
Information exchange 0.4 (0.0-15.2) 0.0 (0.0-23.1) −0.9 (−6.2 to 4.9) 0.0 (0.0-4.2) −2.2 (−7.6 to 0.5) −0.7 (−4.0 to 0.2) 0.0 (0.0-24.8) −0.6 (−5.1 to 8.4) 1.8 (−0.9 to 9.8)
Instruct 1.9 (0.0-31.6) 2.8 (0.0-26.3) −1.3 (−12.9 to 5.1) 3.2 (0.0-10.1) −0.5 (−10.3 to 2.5) 0.6 (−9.5 to 3.9) 3.1 (0.2-11.5) −2.0 (−13.4 to 4.5) −0.7 (−3.8 to 3.8)
Strict instruction 65.2 (0.0-100.0) 37.3 (0.0-85.1) −20.0 (−65.2 to 26.5) 0.0 (0.0-35.5) −24.2 (−80.1 to 35.8) 0.9 (−42.9 to 46.3) 20.5 (0.0-100.0) −15.5 (−77.3 to 39.1) 5.4 (−45.0 to 50.1)
Giving hints 8.2 (0.0-100.0) 37.6 (0.0-100.0) 29.4 (−19.9 to 59.2) 29.8 (0.0-100.0) 38.7 (0.8 to 76.0) 16.4 (−30.4 to 65.6) 52.3 (0.0-100.0) 44.6 (3.1 to 77.8) 1.0 (−50.1 to 50.2)
Provide feedback 18.5 (0.0-39.6) 17.4 (7.0-54.0) 2.4 (−14.9 to 19.4) 16.2 (2.4-30.3) −1.6 (−14.4 to 7.9) −3.0 (−15.6 to 5.7) 23.2 (4.2-40.5) 3.4 (−14.1 to 20.6) 3.5 (−5.2 to 16.7)
Not specified communication 60.2 (17.7-97.3) 52.9 (16.3-85.5) −2.1 (−29.7 to 29.8) 60.6 (21.9-87.1) 2.4 (−13.7 to 23.6) 4.6 (−27.1 to 35.6) 55.7 (18.6-89.3) −2.2 (−24.5 to 30.3) −4.7 (−29.6 to 21.3)
No communication 11.1 (0.0-79.8) 10.3 (0.0-51.6) 2.0 (−21.9 to 22.9) 11.2 (0.5-61.3) 2.8 (−7.0 to 16.0) 2.7 (−22.4 to 29.0) 10.2 (0.0-47.8) 2.6 (−22.2 to 15.9) −3.5 (−26.0 to 13.5)
Neuromotor action
Facilitation 9.8 (0.4-50.7) 2.8 (0.0-11.9) 8.3 (−26.3 to −1.1) 0.9 (0.0-8.4) 10.5 (−25.0 to −3.1) −0.9 (−5.2 to 1.0) 0.7 (0.0-8.8) 11.0 (−27.6 to −3.7) −0.2 (−3.6 to 0.8)
Sensory experience 5.8 (0.4-39.5) 1.0 (0.0-22.9) −7.01 (−20.5 to 2.2) 0.2 (0.0-22.3) −6.2 (−20.7 to 1.2) −0.2 (−6.2 to 5.5) 0.6 (0.0-9.6) −7.2 (−21.1 to 0.8) −0.2 (−6.0 to 3.3)
SPMB 22.2 (9.7-49.5) 41.6 (15.7-73.7) 18.6 (−1.3 to 33.3) 32.4 (11.7-63.9) 10.4 (0.0 to 21.7) −6.3 (−24.6 to 12.6) 41.6 (15.6-84.8) 22.2 (0.8 to 39.9) 11.8 (−10.1 to 32.0)
CSPMB, infant is allowed to continue activity 10.2 (0.0-54.7) 46.6 (4.63-70.8) 25.8 (−6.8 to 45.0) 47.3 (18.8-87.3) 28.8 (3.8 to 54.2) 8.0 (−17.0 to 33.4) 30.7 (7.4-78.7) 17.8 (−11.7 to 46.0) −12.7 (−34.8 to 12.2)
CSPMB, flows over into facilitation, sensory or passive 21.8 (0.0-62.7) 9.6 (1.9-37.2) 15.0 (−34.0 to −1.7) 7.2 (0.0-37.0) 17.1 (−28.1 to −6.3) −1.1 (−10.1 to 11.3) 9.3 (1.6-32.3) −14.4 (−28.6 to 0.0) 3.6 (−9.3 to 13.5)
Not specified neuromotor action 2.0 (0.0-8.1) 1.8 (0.0-3.7) −0.3 (−2.7 to 1.3) 1.3 (0.0-5.6) −0.2 (−3.0 to 2.2) 0.0 (−1.6 to 1.9) 1.4 (0.0-7.2) 0.0 (−3.4 to 3.1) 0.3 (−2.0 to 3.0)
Supine 22.2 (0.0-76.9) 1.8 (0.0-65.4) −19.2 (−53.1 to 11.8) 0.0 (0.0-22.9) 30.3 (−59.7 to −1.1) −7.2 (−31.9 to 5.2) 0.0 (0.0-5.4) 31.4 (−60.7 to −2.5) 0.0 (−8.8 to 2.7)
Prone 22.2 (4.2-72.9) 11.5 (0.0-74.8) 21.5 (−43.7 to −6.1) 14.4 (0.0-32.1) −8.7 (−28.5 to −1.6) 10.3 (−0.8 to 21.2) 10.3 (0.0-91.0) −7.9 (−32.5 to 14.1) 2.0 (−9.0 to 22.5)
Side 6.1 (0.0-27.2) 1.37 (0.0-18.8) 4.6 (−14.2 to −0.1) 0.0 (0.0-53.2) −5.1 (−15.1 to 5.9) −1.0 (−8.6 to 17.2) 0.0 (0.0-15.1) −6.1 (−14.1 to 0.6) 0.0 (−6.1 to 5.9)
Sitting 12.1 (0.0-88.1) 41.5 (6.7-98.7) 22.2 (−18.7 to 58.8) 39.4 (17.7-100.0) 24.4 (−10.8 to 60.6) 5.2 (−18.9 to 28.5) 35.7 (0.0-100.0) 17.4 (−7.9 to 49.5) −4.8 (−33.1 to 21.4)
Standing 0.0 (0.0-19.6) 0.7 (0.0-82.4) 6.2 (−0.7 to 41.2) 27.4 (0.0-80.9) 20.3 (2.1 to 36.8) 7.3 (−21.7 to 31.7) 37.0 (0.0-62.6) 25.6 (5.2 to 44.5) 5.7 (−25.6 to 26.3)
Walking 0.0 (0.0-0.0) 0.0 (0.0-2.8) 0.0 (0.0 to 0.1) 0.0 (0.0-34.1) 0.4 (0.0 to 7.8) 0.3 (0.0 to 7.8) 0.9 (0.0-41.2) 2.3 (0.0 to 20.6) 1.5 (−5.4 to 15.9)
Transition 5.3 (0.4-12.3) 1.7 (0.7-14.2) −2.0 (−5.6 to 1.8) 1.4 (0.0-6.0) −2.9 (−6.5 to 0.2) −0.4 (−5.3 to 2.1) 1.1 (0.0-4.5) 3.6 (−6.9 to −0.4) −0.2 (−2.6 to 1.8)
Motor activity/play 97.1 (54.9-100.0) 100.0 (54.48-100.0) 4.9 (−5.5 to 17.0) 100.0 (86.1-100.9) 4.8 (−2.6 to 19.3) 0.0 (−3.7 to 15.8) 98.9 (86.8-100.0) 3.3 (−2.8 to 19.6) −1.2 (−5.2 to 4.1)
Dressing 2.3 (0.0-26.9) 0.0 (0.0-7.8) −3.9 (−10.7 to 0.0) 0.0 (0.0-0.4) 4.3 (−13.4 to 0.0) 0.0 (−2.5 to 0.0) 0.0 (0.0-0.9) 4.2 (−13.4 to 0.0) 0.0 (0.0 to 0.0)
Carrying 0.6 (0.0-18.0) 0.0 (0.0-3.2 −0.3 (−7.7 to 0.9) 0.0 (0.0-6.2) −0.5 (−6.7 to 0.4) 0.0 (−1.6 to 0.9) 0.0 (0.0-13.2) −0.4 (−6.5 to 3.8) 0.0 (−0.5 to 3.7)
Feeding 0.0 (0.0-0.0) 0.0 (0.0-3.4) 0.0 (0.0 to 0.0) 0.0 (0.0-13.8) 0.0 (0.0 to 0.2) 0.0 (0.0 to 0.2) 0.0 (0.0-4.8) 0.4 (0.0 to 2.4) 0.2 (0.0 to 2.4)
Bathing 0.0 (0.0-0.0) 0.0 (0.0-37.2) 0.0 (0.0 to 0.0) 0.0 (0.0-0.0) 0.0 (0.0 to 0.0) 0.0 (0.0 to 0.0) 0.0 (0.0-0.0) 0.0 (0.0 to 0.0) 0.0 (0.0 to 0.0)
Abbreviations: CI, confidence interval; CSPMB, challenge to self-produced motor behavior; HL, Hodges-Lehmann estimator; PTT, pediatric physical therapy; SPMB, self-produced motor behavior; T0, 1 month before the start of the COPCA course; T1, 3 months after the first part of the COPCA course; T2, 5 months after the first part of the COPCA course or 1 month after the second part of the COPCA course; T3, 7 months after the first part of the COPCA course or since at least 1 month after the third or last part of the COPCA course.
aThe numbers represent the relative duration of therapeutic actions and situations, in percentage of the total treatment time. Results in italics represent significant differences (P < .01). Note that the actions preceded by cubed numeration are modifiers; the modifiers represent relative percentages of time spent in the main category mentioned above the modifier.

At T0 the relative duration of almost all therapeutic actions had wide variability, for example, time spent on “caregiver training” ranged from 0% to 100% (median 42.5%) and time spent on “CSPMB infant is allowed to continue activity” from 0% to 54.7% of PPT time (median 10.2%). In the category “educational actions toward caregiver,” most time was spent on “not specified educational actions” such as the physical therapist performs facilitation techniques without verbal explanation addressed to the caregiver (median 57.5%); 42.5% (median value) of time was spent on “caregiver training” and no time on “caregiver coaching.” In the main category “communication,” most time was spent on the “not specified” subcategory (ie, “not specified communication”), such as communication with the infant, that occurred 60.2% of the treatment time. Specific communication with the caregiver, for example, “information exchange” (0.4%), “instruct” (1.9%), and “provide feedback” (18.5%), occurred less often. In the main category “neuromotor actions,” most time was spent on “self-produced motor behavior” (SPMB; median 22.2%), followed by “CSPMB flows over into facilitation, sensory or passive motor experience” (ie, “flows over into hands-on techniques”; median 21.8%), “CSPMB infant is allowed to continue activity” (median 10.2%), and “facilitation” (median 9.8%).

Course participation was associated with changes in PPT actions (Table 2 and the Figure). At T0 no time was spent on the typical COPCA educational action “caregiver coaching.” During the course, the time spent on “caregiver coaching” increased: between T0 and T1 by 77.9% and between T0 and T3 by 96.1%. Time spent on “not specified educational actions” and “caregiver training” decreased continuously between T0 to T3 (−46.7% and −45.8%, respectively). No changes occurred in the subcategories of “communication” between T0 and T3. The relative duration of the modifier “giving hints” of the subcategory “instruct” increased between T0 to T2 and T0 to T3 (+38.7% and 44.6%, respectively). The duration of all “hands-on” techniques decreased between T0 and T3: “facilitation” decreased continuously between T0 to T3 (−11.0%) and “CSPMB flows over into hands-on techniques” mostly between T0 and T2 (−17.1%). The “SPMB” increased between T0 and T3 (22.2%). The relative duration of “CSPMB the infant is allowed to continue activity” increased between T0 and T2 (28.8%); thereafter the there was no increase (decrease not significant between T2 and T3, −12.7%). Between T0 and T3, the relative duration of the 3 lying positions “supine,” “prone,” and “side” steadily decreased (eg, “supine” −31.4%), while the upright positions “sitting,” “standing,” and “walking” increased (eg, “standing” +25.6%). The relative duration of “motor activity and play” increased between T0 and T2 (+4.8%), while time spent on “dressing” decreased in this period (−4.3%).

Time spent on selected neuromotor and educational actions (%). Data on the relative duration (%) of time spent in 4 neuromotor and 3 educational actions over the time. The horizontal lines represent ranges, the boxes represent the interquartile ranges, and the vertical bars represent median values; the bold points and stars denote outliners. CG indicates caregiver; CSPMB, challenge to self-produced motor behaviors; CSPMB flows over, challenge to self-produced motor behaviors flows over facilitation, sensory, or passive experience; educ, educational; SPMB, self-produced motor behaviors.

Mothers were involved in the PPT sessions (in 63% of the 60 sessions). In some sessions both caregivers or mothers and siblings were involved (15% of the sessions), while 5% of the sessions involved only fathers. At T0, 5 sessions were observed in which family members had an active role in the treatment session. In 10 sessions, family members had no specific role (ie, they watched the PPT performing the therapy session, or their role was restricted to bringing the infant to therapy). In 4 sessions, caregivers performed handling techniques under guidance of PPT instruction. In 1 session, the caregiver acted together with the therapist, with the caregivers guiding the infant's attention. At T1 and T2, family members had an active role in 14 sessions (ie, they acted together with the therapist, hands-off). The caregiver played with the infant while the therapist observed the caregiver-infant interaction and gave hints and suggestions. At T3, the family members actively participated in all 15 sessions.


PPT actions changed after the first days of the COPCA education. This was true in particular for “caregiver coaching,” which increased and “neuromotor actions,” which changed from hands-on to hands-off guidance. After part I, PPT actions continued to change, but at a slower pace, and more for the “coaching” actions than for the “neuromotor actions.”

Physical Therapy Considerations

Time spent on “caregiver coaching” increased between T0 and T1, and it continued to increase at T2 and stabilized thereafter. Simultaneously, time spent on “not specified educational actions toward the caregivers” and “caregiver training” decreased between T0 and T1, a decrease that slowly continued until T3. These changes in the category “educational actions toward the caregivers” indicate changes in the amount and nature of interaction between physical therapist and family and a role shift of the PPTs from a caregiver trainer to a family coach. This was also reflected in the role of the family members during the treatment session. At T0, the majority of family members had no active role, whereas later, family members were actively involved in most of the PPT sessions. The family members played with the infant or were engaged in daily caregiving routines, while the therapist observed the caregiver-infant interaction and gave guidance. These changes express a step toward family-centered practice. Changes in the category communication confirmed that, over time, the PPTs more often applied the communication strategy of “giving hints” as opposed to “strict instructions.” According to the Theoretical Domains Framework, this reflects changes in the domains “skills” and “intentions” (stability of intentions).16 Comparable changes in therapists' behavior were reported by Thomas and Mackintosh22 when they introduced a new program for fall prevention in the elderly. The changes in the category position reported in the present study presumably is an expression of motor development in growing infants and do not reflect changes in the behavior of the PPTs.

During the course, the application of hands-on techniques decreased and that of hands-off techniques increased. This may mean more encouragement of self-produced motor behavior and trial-and-error learning on the part of the PPTs and reflects changes in the domain “behavior regulation” (breaking habits). The implementation of the novel “neuromotor actions” developed at a slower rate than “coaching.” We observed a minor relapse to previous practices in the “neuromotor actions” at the end of the course. Our study suggests that long-term changes in the category of “neuromotor actions” are more difficult to achieve than changes in other categories. The changes in “neuromotor actions” belong to the domain “behavior regulation.” This contrasts with changes in the category of “educational actions toward the caregivers,” which belong to the domains “professional role” (development of role as a coach). These changes are more easily achieved and have higher sustainability. This difference in behavior presumably is brought about by the fact that “neuromotor actions” mostly have been practiced over years and have developed into well-established behaviors (the so-called strong habits), whereas “educational actions toward the caregivers” are less automated and more environment dependent, thereby creating only weak or moderate habits. The neural substrate of strong habits is located in subcortical structure, making them quite resistant to cognitive processes such as generating alternatives or novel goal-directed intentions. Weaker habits are more under explicit cognitive and cortical control, and may be easier to change by goal-directed intentions.23,24 These findings support our hypothesis that changes in the domain of behavior regulation (braking “strong” habits) are the most difficult changes to achieve. To improve the implementation of COPCA in all PPT categories, attention should be given to behavior regulation (ie, more educational support should be provided to change the strong habits involved in neuromotor actions). According to the “behavior change wheel” developed by Michi et al,14 habitual processes belong to the condition “automatic motivation.” Michi et al14 described that possible interventions to overcome behavior change barriers in conditions of “automatic motivation” may consist of environmental restructuring, modeling, and enablement. Also, new skills may develop into automatic responses and new strong habits, if enough repetition and practice is accomplished.22 Hence, possibilities to convert a new behavior into a long-lasting automatic behavior are (1) modeling of the new behavior, (2) sufficient repetition and practice, and (3) a supportive environment.14,22

Methodological Considerations

A major strength of this study is the quantification and determination of the time requirement for changes in time spent on PPT actions and situations throughout the COPCA course, with the valid outcome measure GOP 2.0 as an expression for behavior change. To improve implementation of a new therapeutic approach, understanding the requirements of behavior change is needed. We applied the Theoretical Domains Framework,15,17 a frequently applied basis for empirical studies to explore health professionals' behavior.25,26 The framework allowed us to assign PPT actions to behavior domains and to identify barriers for sustainable behavior changes, a procedure also applied in other studies in the field of physical therapy.24 The additional application of the “behavior change wheel” facilitated the identification of possibilities to overcome behavior change barriers.14 The study has limitations. The last video-recorded PPT session occurred only 2 weeks after the course, precluding knowledge on how well COPCA actions are retained over time. Another limitation is that the video recordings were analyzed by only 1 rater. Masking to the chronology of recording time of the first video was not possible, because it was an easy guess, that the nonhome environment was the precourse PPT session. In addition, we would like to underline that the study demonstrated that participation in a COPCA course was associated with changes in time spent on PPT actions and situations; it did, however, not address the question whether the behavioral changes of the physical therapist were associated with an improvement of infant motor performance. A final limitation is the interdependency of the measurements.


The present study illustrated that assessment of changes in time spent on pediatric physical therapy actions is a suitable method to monitor behavior changes and may promote the implementation of new approaches. The study indicated that after 2 days of education on COPCA PPTs already substantially changed their PPT-actions. The shifts from child to family-centered forms of early intervention and from applying hands-on techniques to hands-off techniques represent successful changes in different domains of behavior (eg, knowledge, skills, professional role, goals, and environment). Changes in the category “educational actions toward the caregivers” were stronger and more sustainable than in the category “neuromotor actions.” The latter represents the difficulty to change “nature of behavior” like strong habits.

To promote changes in the category of “neuromotor actions” in order to achieve long lasting automatic behavior also for this behavior, the following adaptations of the COPCA course will be made: (1) Increase the number of course days (3 × 2 days instead of 1 × 2 days and 2 × 1 day), allowing for more peer feedback; (2) Provide more feedback in the intervals, including peer feedback, as this facilitates appropriate practice and repetition; (3) Pay more attention to the education of the “neuromotor actions”; and (4) Show more videos of optimal COPCA sessions (modeling).


The authors would like to thank all infants and parents who participated in the study as well as the participating therapists and the feedback from Tjitske Hielkema, Elena Mitteregger, Patrizia Sonderer, and Markus Wirz.


1. Dirks T, Blauw-Hospers CH, Hulshof LJ, Hadders-Algra M. Differences between the family-centered “COPCA” program and traditional infant physical therapy based on neurodevelopmental treatment principles. Phys Ther. 2011;91:1303–1322. doi:10.2522/ptj.20100207.
2. Blauw-Hospers CH, Dirks T, Hulshof J, Bos AF, Hadders-Algra M. Pediatric physical therapy in infancy: from nightmare to dream? A two-arm randomized trial. Phys Ther. 2011;91:132–138. doi:10.2522/ptj.20100205.
3. Hielkema T, Blauw-Hospers CH, Dirks T, Drijver-Messelink M, Bos AF, Hadders-Algra M. Does physiotherapeutic intervention affect motor outcome in high-risk infants? An approach combining a randomized controlled trial and process evaluation. Dev Med Child Neurol. 2011;53:8–15. doi:10.1111/j.1469-8749.2010.03876.x.
4. Dirks T, Hadders-Algra M. The role of the family in intervention of infants at high risk of cerebral palsy: a systematic analysis. Dev Med Child Neurol. 2011;53(suppl 4):62–67. doi:10.1111/j.1469-8749.2011.04067.x.
5. Hielkema T, Hamer EG, Reinders-Messelink HA, et al. LEARN 2 MOVE 0-2years: effects of a new intervention program in infants at very high risk for cerebral palsy: a randomized controlled trial. BMC Pediatr. 2010;10:76. doi:10.1186/1471-2431-10-76.
6. Dunst C, Trivette C, Deal A. Enabling and Empowering Families—Principles and Guidelines for Practice. Cambridge, MA: Brookline Books, Inc; 1988.
7. Rosenbaum P. Families and service providers: forcing effective connections, and why it matters. In: Scrutton D, Damiano D, Mayston M, eds. Management of the Motor Disorders of Children With Cerebral Palsy. Cambridge, England: Cambridge University Press; 2004:22–31.
8. Hadders-Algra M. The neuronal group selection theory: a framework to explain variation in normal development. Dev Med Child Neurol. 2000;42(8):566–572.
9. Hadders-Algra M. The neuronal group selection theory: promising principles for understanding and treating developmental motor disorders. Dev Med Child Neurol. 2000;42(10):707–715.
10. Hadders-Algra M. Variation and variability: keywords in human motor development. Phys Ther. 2010;90:1823–1837. doi:10.2522/ptj.2010006.
11. Ives Y. What is “coaching”? An exploration of conflicting paradigms. Int J Evid Based Coach Mentor. 2008;6(2):100–113.
12. Dunst CJ, Boyd K, Trivette CM, Hamby DW. Family-orientated program models and professional helpgiving practices. Fam Rela. 2002;51:221–229. doi:10.1111/j.1741-3729.20002.00221.x.
13. King S, Teplicky R, King GA, Rosenbaum P. Family-centered service for children with cerebral palsy and their families: a review of the literature. Semin Pediatr Neurol. 2004;11:78–86.
14. Michie S, van Stralen MM, West R. The behaviour change where: a new method for characterizing and designing behaviour change interventions. Implement Sci. 2011;6:42. doi:10.1186/1748-5908-6-42.
15. Michie S, Johnston M, Abraham C, Lawton R, Parker D, Walker A. Making psychological theory useful for implementing evidence based practice: a consensus approach. Qual Saf Health Care. 2005;14:26–33. doi:10.1136/qshc.2004.011155.
16. Craig P, Diepppe P, Macintyre S, Michie S, Nazareth I, Petticrew M. Developing and evaluation complex interventions: the new Medical Research Council guidance. BMJ. 2008;337:a1655. doi:10.1136/bmj.a1655.
17. Cane J, O'Connor D, Michi S. Validation of the theoretical domains framework for use in behaviour change and implementation research. Implement Sci. 2012;7:37. doi:10.1186/1748-5908-7-37.
18. Sonderer P, Akhbari Ziegler S, Gressbach Oertle B, Meichtry A, Hadders-Algra M. Psychometric properties of a standardized observation protocol to quantify pediatric physical therapy actions. Pediatr Phys Ther. 2017;29(3):244–250. doi:10.1097/PEP.0000000000000398.
19. Hielkema T, Toonen RF, Hooijsma SJ, et al. Changes over time of contents of infant physical therapy: a quantitative, observation study [published online ahead of print December 21, 2017]. Phys Occup Ther Pediatr. doi:10.1080/01942638.2017.1405863.
20. Shrout PE, Fleiss JL. Interclass correlations: uses in assessing rater reliability. Psychol Bull. 1979:86:420–428.
21. Bauer DF. Constructing confidence sets using rank statistics. J Am Stat Assoc. 1972;67:687–690.
22. Thomas S, Mackintosh S. Use of the theoretical domains of freamework to develop an intervention to improve physical therapist management of the risk of fall after discharge. Phys Ther. 2014;94:1660–1675. doi:10.2522/ptj.2013412.
23. Quellett JA, Wood W. Habits and intention in everyday life: The multiple processes by which past behaviour predicts future behaviours. Psychol Bull. 1998;124:57–74.
24. Savalia T, Shukla A, Bapi RS. A unified theoretical framework for cognitive sequencing. Front Psycol. 2016;7:1821. doi:10.3389/fpsyg.2016.01821.
25. Francis JJ, O'Conner D, Curran J. Theories of behaviour change synthesis into a set of theoretical groupings: introducing a thematic series on the theoretical domains framework. Implement Sci. 2012;7:35. doi:10.1186/174855908-7-37.
26. Little EA, Presseau J, Eccles MP. Understanding effects in reviews of implementation interventions using the theoretical domains of freamework. Implement Sci. 2015;10:90. doi:10.1186/s13012-015-0280-7.

coaching; education; infant; motor development; pediatric physical therapy

Supplemental Digital Content

© 2018 Academy of Pediatric Physical Therapy of the American Physical Therapy Association