The screening Hand Assessment for Infants: Feasibility, validity, and reliability based on parents' recordings

Abstract

Aim

To evaluate the concurrent validity and interrater reliability of remote screening Hand Assessment for Infants (s‐HAI) in infants aged 3.5 to 12 months at risk of unilateral cerebral palsy (CP), and to assess the feasibility of parents administering and recording the s‐HAI play session using a smartphone app.

Method

In this observational study, 30 infants (10 females) at risk of unilateral CP were assessed with remote s‐HAI, administered by their parents, and compared with the HAI conducted in a clinical setting. The quality of the s‐HAI recordings was assessed and parents' experiences were collected using a questionnaire. The Spearman's rank correlation coefficient (r _s ) was calculated for the Each Hand Sum score of s‐HAI and the corresponding six‐item HAI. The interrater reliability of the s‐HAI recordings was calculated using the intraclass correlation coefficient_2,1.

Results

Parents recorded the s‐HAI play session with good quality and found it feasible. A strong correlation of r _s  = 0.86 (p < 0.001) was found between the s‐HAI and the six‐item HAI. Interrater reliability was 0.96 (95% confidence interval [CI] = 0.92–0.99) for the affected hand, 0.95 (95% CI = 0.90–0.98) for the non‐affected hand, and moderate to excellent for individual items.

Interpretation

The s‐HAI measures the same construct as the six‐item HAI, demonstrating reliable scoring between raters when administered remotely by parents, making the s‐HAI a viable tool for the early identification of infants at risk of unilateral CP.

This original article is commented by Boonzaaijer on pages 1243–1244 of this issue.

Plain language summary: https://onlinelibrary.wiley.com/doi/10.1111/dmcn.16378

Abbreviations

EaHS

Each Hand Sum score

HAI

Hand Assessment for Infants

ICC

intraclass correlation coefficient

s‐HAI

screening Hand Assessment for Infants

What this paper adds

• Parents conducted and recorded the screening Hand Assessment for Infants (s‐HAI) with good quality.

• The s‐HAI measured the same construct as the six‐item HAI.

• The scorings of the s‐HAI demonstrated excellent interrater reliability.

• Parents found the s‐HAI procedure feasible to perform.

In recent years, because of increasing evidence supporting early intervention, ¹ there has been a growing effort to identify infants at high risk of cerebral palsy (CP). ² Unilateral CP, a subtype of CP, affecting approximately 35% to 40% of diagnosed children, ³ , ⁴ results from an asymmetric brain injury of various origins, with one early sign being asymmetric hand use. ⁵ Although parents often notice it during infancy, a formal diagnosis typically occurs between 18 and 24 months. ⁶ , ⁷ However, the implementation of a multimodal knowledge translation strategy for early identification can reduce the time to diagnosis. ⁸ , ⁹ This highlights the importance of developing and using standardized assessments specifically designed to detect unilateral CP for the early identification of children at high risk, thereby providing the opportunity for early intervention.

For the early detection of CP, international guidelines ² recommend combining neonatal brain magnetic resonance imaging (MRI) with standardized assessments such as the Prechtl General Movement Assessment ¹⁰ and the Hammersmith Infant Neurological Examination. ¹¹ Among these, the Hand Assessment for Infants (HAI) is particularly notable as the only assessment specifically developed to identify infants at high risk of unilateral CP, focusing specifically on hand use. ¹² With strong predictive validity, ⁵ , ¹³ the HAI serves as an important complement to other assessments for the early identification of unilateral CP; when used together with brain MRI, its predictive value increases. ² , ¹⁴ A recent study suggested that a subset of unilateral HAI items could be used for early screening in infants aged 3.5 to 8 months who are at risk of unilateral CP. ¹⁵ This led to the development of a screening version of the HAI (s‐HAI). The six‐item set for the s‐HAI has shown high accuracy (0.86) in identifying infants at risk of unilateral CP, with good sensitivity and specificity. Pilot testing indicated that the HAI and s‐HAI share a similar theoretical construct, that is, they measure goal‐directed manual actions in infants through a standardized semi‐structured play session. However, prior analysis was based on previously collected HAI data. This study seeks to further investigate the concurrent validity and interrater reliability of the s‐HAI six‐item set in a new data set.

The increasing interest in digital solutions in health care has the potential to transform the delivery of care. ¹⁶ Digitization can increase accessibility, promote equitable person‐centred care, and enhance interventions in a cost‐effective and efficient manner. In paediatric health care settings, parents have a crucial role. They have been involved in data collection for standardized assessments such as the Prechtl General Movement Assessment ¹⁰ via several apps, which have shown high parental satisfaction and good quality of recorded assessments. ¹⁷ , ¹⁸ , ¹⁹ Building on this foundation, we aimed to explore the feasibility of using the s‐HAI remotely. Given that recording the s‐HAI requires more advanced instructions and procedures than previous assessments, further investigation is necessary. Therefore, this study evaluated the concurrent validity and interrater reliability of remotely collected s‐HAI assessments in infants aged 3.5 to 12 months at high risk of unilateral CP, and the feasibility of parents' administering and recording the s‐HAI play session using the NeuroMotion™ eHealth app.

METHOD

Participants and recruitment

In this observational study, a convenience sample of 30 infants was recruited from the follow‐up programme at the neonatal and child neurology department at the Astrid Lindgren Children's Hospital, Karolinska University Hospital, Stockholm, Sweden, from 2021 to 2023. The inclusion criteria were (1) infants aged 3.5 to 12 months (corrected age); (2) with a high risk of unilateral CP because of factors such as a history of a neonatal event such as perinatal stroke (≥ 37 weeks gestational age), or an asymmetric brain injury confirmed using MRI typical for infants born preterm (< 37 weeks), or the presence of observed neurological signs indicating hand asymmetry of unknown origin. Exclusion criteria included infants with unstable health (e.g. uncontrolled seizures) and parents who were unable to communicate in either Swedish or English.

All parents of infants received oral and written information about the study before giving written informed consent. The study was approved by the Regional Ethical Review Board in Lund (no. 2021–01415, amendments 2021–03091 and 2024–04354‐02).

Procedure

HAI recordings were collected as part of a clinical routine by the first author, a clinically experienced occupational therapist. At least one parent participated in the clinical visit. Families were conveniently asked to conduct the s‐HAI play session at home, either before or after their hospital visit, using the NeuroMotion™ eHealth app, ¹⁹ available on the Apple App Store and Google Play. After providing personal login credentials, parents gained access to two instructional videos available in English and Swedish: one guiding them on how to record the play session and one on how to present the toys. The recordings were uploaded to a secure server at Linköping University. All s‐HAI recordings were reviewed for quality; if the recording was not approved, parents were asked to submit a new one. Then, parents answered a digital questionnaire about their experiences of administering and recording the s‐HAI. All except one parent completed the questionnaire. All HAI and s‐HAI recordings were analysed by the first author with approximately 7 to 14 days in between the analyses. For the HAI, only the six unimanual items comparable to the item set for the s‐HAI were scored, hereafter referred to as the six‐item HAI. After initial analysis and feedback from the first eight parental questionnaires, the instructional videos were clarified, particularly emphasizing that the toys should be presented to the side of the infant's body repeatedly. To evaluate interrater reliability in a new data set of remotely collected s‐HAI, the first 16 s‐HAI recordings performed by parents were scored independently by four occupational therapists, all trained in the HAI and with extensive experience in using the assessment.

Assessments

Hand Assessment for Infants

The HAI was developed to evaluate treatment effectiveness ¹² and has shown high accuracy in predicting unilateral CP in infants at risk from 3.5 months of age. ⁵ , ¹³ , ¹⁴ The HAI consists of a semi‐structured, video‐recorded play session lasting 10 to 15 minutes; ¹² it measures goal‐directed manual actions with excellent reliability (intraclass correlation coefficient [ICC] = 0.99, 95% confidence interval [CI] = 0.98–0.99). ²⁰ The HAI is administered by health care professionals who present around 15 carefully selected toys at predetermined locations and positions. ¹² The HAI consists of 12 unimanual items, measuring each hand separately (Each Hand Sum score [EaHS]), and five bimanual items scored on a 3‐point rating scale. An asymmetry index of hand function is calculated based on the EaHS (Table 1).

TABLE 1.

Test characteristics of the screening Hand Assessment for Infants and Hand Assessment for Infants.

Characteristic	s‐HAI	HAI
Measurement properties	Predictive	Descriptive
		Evaluative
		Predictive
Age focus	3.5–12 months	3.5–12 months
Number of items	6	17
Unimanual	6	12
Bimanual	0	5
Total raw score EaHS, range	0–12	0–24
Setting	Home environment and clinic	Clinic
Number of toys to use	5–6	15
Length of video‐recorded play session	5 minutes	10–15 minutes
Time to score the assessment	About 10 minutes	About 30 minutes
Assessors	Guardian or professional after looking at instructional videos	Trained professionals
Course	Self‐administered online course for professionals, 10–15 hours	2.5 days for professionals

Abbreviations: EaHS, Each Hand Sum score; HAI, Hand Assessment for Infants; s‐HAI, screening Hand Assessment for Infants.

Screening Hand Assessment for Infants

The s‐HAI is an assessment designed to identify infants at high risk of developing unilateral CP from 3.5 months of age. ¹⁵ Incorporating elements from the HAI, the s‐HAI aims to share the same theoretical concepts and procedure as the HAI, but it is only used in a 5‐minute play session including five to six carefully selected toys. The s‐HAI assesses six unimanual items (Table 2) scored on a 3‐point rating scale, using the same scoring criteria as the HAI. In this study, we investigated for the first time the s‐HAI when remotely collected by the infants' guardians. Further differences and similarities between the HAI and s‐HAI are described in Table 1.

TABLE 2.

The six‐item set for the screening Hand Assessment for Infants based on the unimanual items from the Hand Assessment for Infants.

Item number from the HAI	Item name
1	Initiates hand use for objects on the side
4	Quality of holding
5	Grasps from an easy position
6	Object location when grasping
8	Moves fingers
11	Quality of hand and arm movements

Abbreviation: HAI, Hand Assessment for Infants.

Criteria for video quality

To assess the necessary quality of the s‐HAI recordings, a form with specific criteria was developed. The form consists of three domains: recording (one item); infant setting (four items); and play session (three items) (Table 3). To approve an s‐HAI recording, specific criteria had to be met in each domain: (1) the recording had to be graded as acceptable; (2) in infant setting, three of four items had to be graded as acceptable; and (3) in play session, two of three items had to be graded as acceptable. If any domain did not meet approval, a new recording was requested.

TABLE 3.

Description of the criteria for video quality, and video quality for recordings of the screening Hand Assessment for Infants used in this study (n = 30).

Domain	Item	Acceptable	Not acceptable
Recording	Is the camera positioned in front of the child?	30	0
Infant setting	Is the child as upright as possible?	30	0
	Are the arms able to move freely?	29	1
	Are the hands and arms visible?	30	0
	Is the environment during the recording distracting?	29	1
Play session	Is the child in a fairly good mood?	30	0
	Is the selection of toys appropriate?	30	0
	Are the toys offered from varying positions?	27	3

Parents' experiences of performing the screening Hand Assessment for Infants

A digital questionnaire was developed to investigate parents' experiences of administering and recording the s‐HAI play session, and to what extent the instructions for performing the play session were possible to follow. There were 10 questions, including eight with multiple‐choice options and two with commentary fields. The questionnaire was sent by mail to the parents after approval of the recording.

Statistical analyses

Both the s‐HAI and six‐item HAI are scored on a 0 to 2‐point rating scale with a sum score ranging from 0 to 12 for each hand (EaHS). Analyses for interrater reliability and concurrent validity were performed with SPSS v28.0 (IBM Corp., Armonk, NY, USA).

Concurrent validity was investigated using the Spearman's rank correlation coefficient (r _s ). Interpretation of the r _s values was categorized as follows: small (0–0.30); low (0.30–0.50); moderate (0.50–0.70); high (0.70–0.90); and very high (0.90–1.00). ²¹

Interrater reliability was calculated using the ICC_2,1, two‐way random effects model (absolute agreement, average measures), with 95% CI. The ICC was calculated for the EaHS and individually for each item for both the infant's affected and non‐affected hands. The required sample size of raters and s‐HAI recordings was calculated according to the methods outlined by Zou ²² in R v4.0.5 (R Foundation for Statistical Computing, Vienna, Austria), with the R package ICC.Sample.Size v1.0. We aimed for quality parameters for the ICC values between 0.725 and 0.9, with alpha = 0.05 and power set to 80%. Interpretation of the ICC was as follows: poor (<0.5); moderate (0.5–0.75); good (0.75–0.9); and excellent (>0.9). ²³ Therefore, four raters had to independently assess the same 16 s‐HAI recordings.

RESULTS

Participants

A total of 30 infants, 10 females and 20 males, aged 3.5 to 12 months corrected age were included. The infants' ages varied depending on when they were referred to the occupational therapy department. The assessments of the s‐HAI and HAI were conducted with an average interval of 5.4 (range: 0–12) days between them. In 10 convenience cases, the s‐HAI was conducted before the HAI assessment, while in 20 cases it was conducted after. The EaHS for both the s‐HAI and six‐item HAI ranged from 0 to 12 for the affected hand, 2 to 12 for the non‐affected hand for the s‐HAI, and 5 to 12 for the six‐item HAI. The characteristics of the infants and their parents are presented in Table 4.

TABLE 4.

Demographic data of participants (n = 30) and their parents (n = 30).

Characteristic	Total (n = 30)
Gestational age at birth (weeks), median (25–75th centile)	37.5 (30.8–40)
Females/males, n	10/20
Born at term/preterm (< 37 weeks' gestation) a	15/15
Age at the s‐HAI assessment (months corrected age), median (25–75th centile)	5 (3.8–7.3)
Age at the HAI assessment (months corrected age), median (25–75th centile)	5 (4–8)
EaHS for the s‐HAI, median (25–75th centile)
Affected hand	8 (5–9), range 0–12
Non‐affected hand	10 (8–11), range 2–12
EaHS for the six‐item HAI, median (25–75th centile)
Affected hand	8 (7–10), range 0–12
Non‐affected hand	10 (8–12), range 5–12
Inclusion category
Clinical diagnosis of stroke	13
Asymmetrical perinatal brain injury verified using MRI	8
Neurological signs of hand asymmetry	5
Asphyxia	4
s‐HAI recording made by
Mother	21
Father	3
Both parents	5
Missing data	1
Parents' age, years
20–29	4
30–39	23
> 40	2
Missing data	1
Parents' education
Elementary school	0
High school	7
Post‐secondary education	3
College or university education	19
Missing data	1

^a Term: ≥ 37 weeks gestational age at birth; preterm: 30–36 weeks gestational age; very preterm: < 30 weeks gestational age. Abbreviations: EaHS, Each Hand Sum score; HAI, Hand Assessment for Infants; s‐HAI, screening Hand Assessment for Infants.

Video quality of parents' s‐HAI recordings

Of the 30 s‐HAI recordings, 28 were deemed acceptable in the domain recording, while two required re‐recording because of lack of quality and technical issues; one recording was conducted before the clinic visit, and the other after it. Two families submitted more than one recording; in these cases, the recording with the highest quality was included. Although some recordings did not meet all the criteria for the specific domains, they were approved according to the set standards of video quality developed for this study. Specifically, in the infant setting domain, two recordings had one unacceptable item each because of the infants' hands and arms not being consistently visible, and because of environmental distractions. However, as three out of four items were acceptable, these recordings were approved. In the play session domain, three recordings had one item marked as unacceptable, primarily because toys were not presented from varying positions (Table 3).

Parents' experiences of performing the s‐HAI

The digital questionnaire was answered by 29 of 30 parents (Table 5). Most parents (n = 24) found the instructions on how to perform the play session and how to present the toys sufficient (n = 25). Most found it pretty easy or easy (n = 27) to conduct the play session and to find suitable toys (n = 23). One parent experienced difficulties in both administering the play session and finding suitable toys; this parent also recorded the session more than twice before submission. Most parents (n = 22) observed that their infant used their hands as usual during the play session. Recording the play session several times was optional before submission; nearly half of the parents (n = 14) used that option. However, only two parents submitted more than one recording. The reasons for recording multiple times were not documented.

TABLE 5.

Questionnaire of parents' experiences of recording the play sessions for the screening Hand Assessment for Infants (n = 30 parents).

Question	Possible response options	Answers (n = 29)
1. Were the instructions on how to arrange the recording sufficient?	Yes, sufficient	24
	Partly sufficient	5
	Insufficient	0
2. Were the instructions on how to present the toys sufficient?	Yes, sufficient	25
	Partly sufficient	4
	Insufficient	0
3. What was it like to perform the play session?	Easy	10
	Pretty easy	17
	Difficult	2
	Very difficult	0
4. What was it like to find toys that were similar to those in the instructional video?	Easy	8
	Pretty easy	15
	Difficult	6
	Very difficult	0
5. How do you think your child uses the hands during the recording compared to how they usually do?	More than usual	0
	As usual	22
	Less than usual	7
	Much less than usual	0
6. How satisfied are you with the recording you submitted?	Very satisfied	3
	Satisfied	24
	Dissatisfied	2
	Very dissatisfied	0
7. Did you have to redo the recording?	No	15
	Once	10
	Twice	2
	More than twice	2
8. Do you prefer to record it yourself at home or at the hospital?	At home	16
	At the hospital	6
	It does not matter	7

More than half of the parents (n = 16) preferred to self‐administer and record the s‐HAI play session remotely, while seven parents had no preference regarding self‐administration. Those (n = 6) who preferred to have the s‐HAI performed by an occupational therapist at the hospital cited reasons such as the complexity of the play session, lack of suitable seating, or preference for professional assistance during the recording process. Additionally, two parents felt they missed the opportunity to consult a professional when conducting the assessment at home.

Concurrent validity of s‐HAI

The correlation between the EaHS of the s‐HAI and six‐item HAI was r _s  = 0.86 (p < 0.001) for both the affected and non‐affected hands. There was no correlation between age and the EaHS for the affected hand (r _s  = 0.071, p > 0.5) in the s‐HAI or the six‐item HAI (r _s  = 0.07, p > 0.5). This indicates that the affected hand score was not age‐dependent, either for the s‐HAI or the six‐item HAI.

Interrater reliability

The interrater reliability of the s‐HAI was excellent for the EaHS of the affected (ICC = 0.96, 95% CI = 0.92–0.99) and non‐affected (ICC = 0.95, 95% CI = 0.90–0.98) hands (Table 6). For individual items, the ICCs for the affected hand were excellent for items 1 and 5, and good for items 4, 6, 8, and 11. For the non‐affected hand, the ICCs were excellent for item 4; good for items 5, 6, and 11; and moderate for items 1 and 8 (Table 7).

TABLE 6.

Interrater reliability of the Each Hand Sum score of the screening Hand Assessment for Infants between four experienced raters for the affected and the non‐affected hands (n = 16).

EaHS	ICC (95% CI)
Affected hand	0.96 (0.92–0.99)
Non‐affected hand	0.95 (0.90–0.98)

Abbreviations: CI, confidence interval; EaHS, Each Hand Sum score; ICC, intraclass correlation coefficient.

TABLE 7.

Interrater reliability of individual items for the screening Hand Assessment for Infants between four experienced raters for the affected and non‐affected hands (n = 16).

Item	ICC (95% CI) affected hand (n = 16)	ICC (95% CI) non‐affected hand (n = 16)
Item 1 (initiates hand use for objects on the side)	0.92 (0.83–0.97)	0.73 (0.44–0.89)
Item 4 (quality of holding)	0.89 (0.77–0.96)	0.94 (0.88–0.98)
Item 5 (grasps from an easy position)	0.94 (0.86–0.98)	0.85 (0.69–0.94)
Item 6 (object location when grasping)	0.86 (0.71–0.95)	0.84 (0.67–0.94)
Item 8 (moves fingers)	0.80 (0.58–0.92)	0.65 (0.29–0.86)
Item 11 (quality of hand and arm movements)	0.89 (0.76–0.96)	0.83 (0.65–0.93)

Abbreviations: CI, confidence interval; ICC, intraclass correlation coefficient.

DISCUSSION

Based on the results from this study, along with previous research on the development of the s‐HAI, ¹⁵ the s‐HAI emerges as a viable tool for screening infants at high risk of unilateral CP. The s‐HAI demonstrates good validity, reliability, and feasibility when administered remotely by parents to infants at high risk of unilateral CP. Validity is evidenced by the nearly equivalent scoring results between the s‐HAI and HAI for the six comparable items. This suggests that a remotely collected s‐HAI and the corresponding six items in the HAI reflect the same theoretical construct. The interrater reliability of the s‐HAI was excellent for the EaHS and moderate to excellent for the individual items when scored by experienced raters.

The procedure for recording, including the instructional videos developed in a previous study, ¹⁵ required only minor adjustments based on the initial analysis of the s‐HAI recordings and parental feedback. Parents were able to administer and record the s‐HAI play session with good quality after viewing two short instructional videos. This aligns with findings from other studies where parents remotely recorded the Prechtl General Movement Assessment ¹⁷ , ¹⁸ , ¹⁹ and the Alberta Infant Motor Scale. ²⁴ Although the Alberta Infant Motor Scale involves a more complex data collection process and parents found the assessment somewhat complicated and time‐consuming to perform, ²⁵ , ²⁶ they still preferred conducting the assessment at home despite the difficulties in finding suitable times to collect the data. Similar time constraints were mentioned by some parents in this study, although most felt that they saved time by not needing to visit the hospital. Little is known about the impact on parents when involved in data collection. Previous studies reported some increased insecurity about their child's motor development, but also that the experience helped them gain a deeper knowledge and awareness of their child's capability, making the recordings meaningful. ¹⁹ , ²⁵ In this study, we explored technical and practical aspects of administering and recording the s‐HAI via an app. Further research is needed to explore parents' experiences of assessing their own child with the s‐HAI.

For the development of the s‐HAI, the original HAI, often considered the criterion standard, was used as an external criterion. The strong correlation between the same six items from the HAI and shortened HAI recordings in a previous study (r _s  = 0.91), ¹⁵ along with the correlation between the six‐item HAI and s‐HAI (r _s  = 0.86, p < 0.001), strengthens the concurrent validity of the s‐HAI. Interrater reliability for the EaHS was excellent, suggesting that the s‐HAI can be used at an individual level for decision‐making and diagnostic support. Guidelines recommend a reliability coefficient of at least 0.90 for individual‐level decision‐making, ²⁷ although a coefficient between 0.80 and 0.90 can be acceptable depending on the purpose of the measurement. ²⁸ , ²⁹

The ICC values for the EaHS of the s‐HAI were slightly lower in this study compared to the EaHS of the HAI (ICC = 0.99, 95% CI = 0.98–0.99) reported in a previous study. ²⁰ This discrepancy may be due to the smaller sample size for the s‐HAI (s‐HAI n = 30, HAI n = 50) and lower variance. ³⁰ Additionally, a measurement scale with more items typically increases the variance and affects the ICC, ³¹ which might explain the difference for the HAI. In this study, we used a model with four experienced raters to increase variation and precision; having more raters generally leads to greater variance and accuracy. ³¹ Despite the overall excellent ICC for the EaHS, two individual items only achieved a moderate ICC for the non‐affected hand (Table 7).

For item 1 (initiates hand use for objects on the side), assessors noted that several parents did not consistently present toys to the side of the infant's assessed hand according to the manual. This inconsistency may have led to different scoring interpretations by the assessors. These instructions were also part of the clarifications made to the instructional videos. This issue might have been avoided if we had used the s‐HAI recordings collected later in the study for reliability testing, as most of the participants used the updated instructional videos. Item 8 (moves fingers) also had moderate interrater reliability for the non‐affected hand. This item typically receives full marks for typically developing children as early as 3 months of age; it is assumed that children with unilateral CP have one typically developing hand. However, in this study, some infants, given their risk factors, may have had involvement of the non‐affected hand. Although the assessors in this study have extensive experience with the HAI, they might have missed minor deviations or disagreed with the scoring.

Methodological considerations

Data collection occurred in different sequences within the sample, with 10 out of 30 s‐HAI recordings completed before the HAI in the clinic. It is likely that parents who recorded the s‐HAI after the clinic visit gained additional insight into the execution of the play session. However, we do not know whether it was the same parent who recorded the s‐HAI who also attended the clinical visit. Notably, in two parents who needed to submit a new recording because of quality issues, one recorded the s‐HAI before and the other after the clinic visit.

Another limitation of this study is that the reason for recording the s‐HAI multiple times before submission is unknown. There were no specific comments or remarks in the parental questionnaire and we found no clear relationship between the responses regarding recording the s‐HAI more than once (Table 5, question 7) and whether parents found the play session difficult to perform (Table 5, question 3), except for one parent. Possible reasons for re‐recording may include the infant's mood, parents forgetting to use some toys, or uncertainty about the quality of the recording. As the s‐HAI is designed to be a playful and harmless play session for the infant, repeating it does not pose any risks. However, parents' experiences of administering an assessment for their own child warrants further investigation.

A further limitation of this study is that the first author scored both the s‐HAI and the six‐item HAI assessments. However, given the workload of scoring other children in the clinic, it is unlikely that specific scores for individual children could be recalled after a gap of about 2 weeks.

A convenience sample has inherent limitations, such as potential motivation bias, and the results may be strictly applicable only to the studied population. However, the sample in this study showed variability in infant characteristics and hand function (Table 4), suggesting that the sample is fairly representative of infants at risk of unilateral CP. The socioeconomic status of the families may be a limitation as only families speaking Swedish and English and owning a smartphone were included. Additionally, most parents (n = 19) had a college or university degree, and all had education beyond elementary school. A previous Australian study found that maternal education and proficiency in English had an impact on the execution of the assessment when using an app. ¹⁷ Therefore, further investigation is needed to assess the use of the s‐HAI in different contexts.

Clinical implications

The s‐HAI is a usable tool that can enhance accessibility to care and promote more equitable care among infants at risk. It measures the same construct as the six‐item HAI and can be used effectively in a remote setting to identify infants at high risk of unilateral CP, with moderate to excellent interrater reliability. Parents are capable of successfully administering and recording an s‐HAI play session in a home environment with good quality using an app; parents found the procedure feasible.

CONFLICT OF INTEREST STATEMENT

The authors have no conflict of interest.

Kembe J, Holmström L, Broström E, Sundelin H, Eliasson A‐C. The screening Hand Assessment for Infants: Feasibility, validity, and reliability based on parents' recordings. Dev Med Child Neurol. 2025;67:1301–1308. 10.1111/dmcn.16318

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.