Study reference

Study characteristics

Patient characteristics

Prognostic factor(s)

Follow-up

Outcome measures and effect size

Study quality

Research question: What factors are associated with reaching ambulation in children and adolescents with spina bifida?

Bartonek, 2001

Type of study: cross-sectional study

Setting: hospital based

Country: Sweden

Source of funding:

Grants from Norrbacka Eugenia

Foundation and Foundation ‘Konung Oscar II och Drottning

Sophias Guldbröllopsminne’.

Inclusion criteria: consecutive series of children with myelomeningocele (MMC) born between 1985 and 1992 considered to have reached an assessable ambulatory level

Exclusion criteria: lack of motor deficit (3), unwillingness of parents to participate (5), not living in the area (1)

N=53 alleen inclusion

Mean age ± SD: 7.6 (3.2-11.4)

Sex: % M / % F

53/47

Other important characteristics: 6 children attended a school for learning difficulties

Prognostic factor(s) and method of measurement:

Muscle strength (manual muscle testing – graded 0-5, Daniels, 1995))

Spasticity (Stark &Baker, 1967)

Joint contractures and deformities - of the ankle (10° or more fixed plantar flexion), knee and hip (20° or more from a neutral joint position)

Balance disturbance in standing, walking or sitting

(Nora, 1996)

Hypotonia (generalized muscle hypotonia not related to the flaccid paresis of single muscle groups)

Obesity (BMI)

Hip dislocation (medical reports)

Degree of scoliosis (medical reports)

Orthopaedic and neurosurgical procedures performed (medical reports)

Shunted hydrocephalus

(medical reports)

Shunt revisions

(medical reports)

Orthoses (orthotic use, or use of other ambulatory aids or wheelchair)

Physical effort during walking (physical cost index- PCI) N=39

Functional mobility (PEDI)

Muscle function class (Bartonek, 1999, Sharrard, 1964, Smith&Smith, 1973)

Endpoint of follow-up:

n.a.

For how many participants were no complete outcome data available?

N (%): 53=100%

14/53 geen PCI

Reasons for incomplete outcome data described? Yes

Outcome measures and effect size

31 of 53 children had reached the expected ambulation considered possible according to their motor paresis (group A), whereas 22 of the 53 performed worse than expected (group B). Ambulatory function was assessed by Hoffer criteria (1973).

Spasticity was found in 13 of 31 children in group A (42%) and 13 of 22 children in group B (59%). (p=0.002)

26 of 31 children in group A (84%) and 20 of 22 children in group B (91%) had shunted hydrocephalus. Eleven children in each group had shunt revisions. In 12 children (four from group A and eight from group B) two or more shunt revisions had been performed. This was statistically significant.

balance disturbances sitting, standing, walking

2 of 31 (6.45%) vs 9 of 22 children (40.9%)) sign.

generalized muscle hypotonia

0 of 31 vs 4 of 22 children (18.1%) sign.

D

- Cross-sectional study design (no causal relationships)

Bartonek, 2010

Type of study:

Prospective cohort

Setting:

Hospital based

Country:

Sweden

Source of funding:

Not reported

Inclusion criteria:

all children diagnosed with MMC who were born with spina bifida (SB) between July 1993 and September 1999 and who were treated at the Karolinska University Hospital

Exclusion criteria: meningocele, additional diagnosis of juvenile idiopathic arthritis, not able to participate

N=43

Mean age ± SD: n.a.

Clinical visits were planned at 6 months and at 1, 1.5,

2, 3, 4, and 6 yrs of age

Sex: % M / % F

53/47

Other important characteristics:

40 of the children had shunted hydrocephalus (93%)

Prognostic factor(s) and method of measurement:

Muscle strength (0-5, Daniels, 1995 and in infants by observation)

Muscle Function Class (1-5)

MFC I indication the lowest neurological level of impairment with weakness of intrinsic foot muscles, MFC II en MFC III with weakness or absence, resp., of plantar flexion, knee flexion, hip extension, and abduction activity, MFC IV with weak or no knee extension and hip adduction but good pelvic elevation activity, and MFC V indicating the highest level with no muscle activity in the lower limbs. Compared to Sharrard (1964) resp. S2, S1-L5, L4-L3, L1-L2, Thoracic

Spasticity (absent/present)

Range of motion of lower limb joints (recommendations of American Academy of Orthopedic Surgeons)

Joint contractures of the ankle (10° or more fixed plantar flexion), knee and hip (20° or more from a neutral joint position)

Hip extension rate (Thomas test, Bartlett, 1985)

Hip disclocation (medical records)

Orthopedic surgeries (medical records)

Use of orthotics (prescribed according to guidelines – MFC-group)

Motor development (towards ambulation) (observation)

Hypotonia (generalized muscle hypotonia not related to the flaccid paresis of single muscle groups)

Endpoint of follow-up:

In the 43 children 251 assessments were performed, on an average of 5.84

occasions (range, 3–7). 36 children were examined at app. 6 months (mean0.58 years, SD0.1),

38 at 1 year (1.1, SD 0.2), 39 at 1.5 years (1.7, SD 0.2), 36 at 2 years (2.3, SD 0.2), 28 at 3 years ( 3.1, SD0.3), 36 at 4 years (4.3, SD0.3), and 38 at 6 years (6.8, SD 0.6).

For how many participants were no complete outcome data available?

N (%): 5 (12%)

Reasons for incomplete outcome data described? no

Outcome measures and effect size

Walking function (modified Hoffer criteria, 1973) had been achieved at the 1-year follow-up (FU) in 2 of 38 children, at the 1.5-year FU in 7 of 39, at the 2-year FU in 14 of 36, at the 3-year FU in 21 of 28, at the 4-year FU in 28 of 36, and at the 6-year FU in 30 of 38. At the 6-year FU up 35 of 38 children walked independently with (13) or without a walking aid (17), of whom 5 required human support.

In MFC I (n=1) child walks independently at 1 yr-6 yr.

MFC II 1 (of 23), 6 (of 24), 10 (of 21), 14 (of 16), 17 (of 22), 14 (of 16) walks independently or with walking aid at resp. 1, 1.5, 2, 3,4 and 6 yr of age,

In MFC III 2 (of 8), 4 (of 5), 8 (of 8), 12 (of 15) walks independently or with walking aid at resp. 2, 3, 4 and 6 yr of age.

In MFC IV 1 (of 1), 1 (of 1), 1 (of 2) walks with walking aid at resp. 2, 4 and 6 yr.

IN MFC V 1 (of 5), 1 (of 5), 1 (of 4), 2 (of 4), 2 (of 4) walks with walking aid at resp. 2, 3, 4 and 6 yr of age.

Characteristics of children not having achieved expected ambulatory function at the 6-year follow-up:

- increased incidence of spasticity during the entire follow-up period compared with those who achieved expected ambulatory function (9 of 9 versus 15 of 31; p= 0.006)

-spine surgery performed more frequently (4 of 9 versus 2 of 31; p_0.016)

- increased use of inadequate orthoses (5 of 7 versus 4 of 31; p= 0.004).

- Use of orthoses of for less then 5 hours a day (22 of 30 versus 1 of 7; p = 0.006).

B

-No multivariate analysis has been performed

Biggio, 2001

Type of study:

Retrospective cohort

Setting: hospital based

Country: USA

Source of funding: not reported

Inclusion criteria: all pregnancies complicated by fetal open spina bifida

that were managed prenatally and postnatally at University Hospital of Alabama from January 1996 through March 2000.

Exclusion criteria:

fetuses with other organ system

malformations or aneuploidy and pregnancy terminations (20).

Also excluded three fetuses that underwent in utero closure of spinal defects.

N=33

Mean age ± SD: 21.9 (4.9) – wks (prenatal)

Sex: % M / % F

Not reported

Other important characteristics:

Postnatally all infants underwent standard neurosurgical closure

Prognostic factor(s) and method of measurement

Lesion level (highest vertebral level at which dysraphism was visualized) (ultrasound*)

Lesion type: meningocele, myeolomeningocele, myeoloschisis (ultrasound)

Lateral ventricle width/ ventriculomegaly (ultrasound)

Talipes (club foot) (ultrasound)

Shunt

* Accuson 128 XP (Accuson Corp, Mountain View, Calif) with a 3.5

MHz or 5 MHz probe or a General Electric Logic 700 with

a multi-frequency (4-7 MHz) probe (GE Medical Systems,

Milwaukee, Wis).

Before August 1997, static images were recorded on

thermal photo-paper. Since August 1997, all images

stored in digital archiving system.

Endpoint of follow-up:

16.7 (13.5) mo

Complete outcome data for all children (33)

Outcome measures and effect size

ambulatory or nonambulatory.

 Ambulatory: children ≥2 years old had to walk, with or without appliances, and those <2 years of age at last follow-up had to have at least 4/5 muscle strength in the lower extremities. Muscle power was graded on a scale of 0-5 as follows: 0, no contraction; 1, trace of contraction; 2, active movement with gravity eliminated; 3, active movement against gravity; 4, active movement against gravity and resistance; 5, normal movement. Scoring was based on spontaneous activity, the ability and willingness to support body weight when suspended from the axillae, and testing of individual muscle groups.

Infants with thoracic lesions (n = 11), none were ambulatory, all had ventriculomegaly diagnosed antenatally, and all required postnatal shunt placement.

Nearly half had club foot diagnosed antenatally.

Infants with L1-L3 lesions (n = 12), 50% were ambulatory. There was no difference between the ambulatory and nonambulatory groups in terms of the incidence of ventriculomegaly, ventricular width, antenatal presence of club foot, or need for postnatal shunting.

All infants with L4-sacral lesions (n = 10) were ambulatory, and although only 60% had ventriculomegaly diagnosed antenatally, 90% ultimately required postnatal shunt placement.

More caudal (lower) lesion level (P < .001) and smaller lateral ventricular width seen in the ultrasound examination most proximate to delivery (P = .003) were associated with ambulatory status (univariate analyse). The presence or absence of club foot did not affect the likelihood

of ambulatory status (P = .39).

More caudal lesion level was significantly associated and there was a trend for smaller ventricular size to be associated with ambulatory status, P = .001 and P = .09, respectively (multiple logistic regression to control for any association between lesion level and ventricular size)

B

- Retrospective cohort study

Charney, 1991

Type of study: Retrospective cohort study

Setting:

Hospital based

Country:

USA

Source of funding:

Not reported

Inclusion criteria: All children followed through the age of 5 years by the Spina Bifida Program at the Children’s Hospital of Philadelphia since 1961 who had MMC and high lumbar or thoracic level paralysis and who received bracing by 3 years of age.

Exclusion criteria:

Severe cerebral palsy and upper extremity involvement (n=10), Chiari II respiratory complaints (n=3).

N=87

Mean age: 16 (range 6-28)

Sex: % M / % F

48/52

Other important characteristics:

94% of the children had a shunt

Prognostic factor(s) and method of measurement:

Presence of hydrocephalus

Type of bracing and walking device first decribed

Spinal deformity by 5 yr of age

Orthopedic surgery of the hips by 5 yr of age

Intellect (mentally challenged IQ<70) at age > 5 yr.

Physical therapy received directed at walking

Parent compliance with walking training and wearing braces

All data from medical records.

Endpoint of follow-up: retrospective analysis from birth 16 (6-28) yr, every 5 years, but focus was status at 5 yr

For how many participants were no complete outcome data available?

N (%):

Reasons for incomplete outcome data described?

Outcome measures and effect size:

Community ambulation, household ambulation, exercise ambulation. For purpose of analysis two groups: community ambulators (CAMS) and non-community ambulators (non-CAMS).

Congenital kyphosis (p=0.007), the use of either a standing frame (p=0.005) or rollator walker (p=0.001) were sign. more common among non-CAMS.

Physical therapy directed at walking, compliant parents and being not mentally challenged were sign. more common among CAMS (p<0.001). Multivariate analysis identified the last three factors as the most important in ifferentiating between CAMS and non-CAMS.

52% of children were CAMS by 5 yrs of age.

B

- Retrospective cohort study

Cochrane, 1996

Type of study: 2 retrospective cohort study and retrospective diagnostic accuracy study.

Setting: hospital based

Country: Canada

Source of funding: not reported

Inclusion criteria A: fetuses with spinal dysraphism diagnosed at the Center for Prenatal Diagnosis and Treatment , British Columbia Women’s Hospital between 1987 and 1991 for whom fetal ultrasound and postnatal pathological and radiological examinations had to allow confirmation of the prenatal diagnosis.

Exclusion criteria A:

N=26

Mean age ± SD: 15 < 22 wks; 23 wks < 7 < 30 wks, 4 > 30 wks.

Sex: % M / % F

Not reported

Other important characteristics: prenatal

Inclusion criteria B:children born at the Meningomyelocele Clinic at British Columbia Children’s Hospital with

MMC between 1971-1981 who have been treated aggressively throughout life in a manner identical to current management

Exclusion criteria B:

N=85

Mean age ± SD: (9-19 yr) (in 1990 analysis of children born between 1971-1981)

Sex: % M / % F

Not reported

Other important characteristics:

Prognostic factor(s) and method of measurement:

A.Retrospective diagnostic accuracy study: predicted functional outcome based on prenatal determination of most superior level of dysraphism compared with postnatal pathological or radiographic examinations and actual outcome and expected outcome based op study population B.

B.Level of MMC:

-thoracolumbar (last intact level at L2), lumbar (last intact level at L3-4), sacral (last intact level at L5)

Endpoint of follow-up:

A: less then 5 yrs

B: Follow-up in 1990

Children born between 1971 en 1981 (9-19 yr)

For how many participants were no complete outcome data available?

N (%): n.a.

Reasons for incomplete outcome data described?

Outcome measures and effect size

A.From an eventual functional perspective, the prenatal study accurately localized the cranial level of the dysraphic state in all but the fetus with hemiMMC. Only 3 of the 26 fetuses had a daviation of 3 vetrebral bodies from the predicted to the actual level, but these were above L2 > no functional predictions incorrect.

B.Locomotion (walking without aids, with ankle foot orthoses (AFO) or insert:

inside ambulatory, community ambulatory with braces and crutches or walker, total community ambulators, wheelchair.

10% of 20 thoracolumbar patients (2/20), 52% of 31 lumbar-level (16/31) and 83% of 30 sacral level patients (25/30) became total community ambulators.

B

- Retrospective cohort study

Fraser, 1992

Type of study: retrospective cohort

Setting: hospital based

Country: South Africa

Source of funding: not reported

Inclusion criteria: Patients presenting with mid-lumbar neurological MMC during 1970-1986.

Exclusion criteria: ≤4 yr

N=55

Mean age ± SD: not reported

Sex: % M / % F not reported

Other important characteristics:

Prognostic factor(s) and method of measurement:

Hip stability

Fixed flexion deformities ≥10◦, deformities of spine, ankle or foot.

Spasticity in lower limbs

Intelligence

Neurological lesion level:

L3 quadriceps power of MRC ≥ grade 3 medial hamstrings ≤ grade 3 (n=24)

 L4 :quadriceps power ≤ MRC grade 3, medial hamstrings ≤ grade 3. (n=14)

Radiography of hips, dislocated or subluxated hips were considered to be unstable.

Orthotic use

Surgical procedures

Endpoint of follow-up: 10.2 yr (4-19)

For how many participants were no complete outcome data available?

N (%): 20 (36%)

Reasons for incomplete outcome data described?

Yes (n=7), no (n=13) – 24%

Outcome measures and effect size

Walking ability (Hoffer, 1973) Only community ambulation was considered useful ambulation.

Mean age of walking 3 yr, 8 mo (2 yrs, 6 mo – 5 yr, 6 mo)

L4 patients walked on average six months earlier than L3 patients (3yrs 6 mo as opposed to 4 yrs 1 mo)

All ambulant patients used bilateral elbow crutches and below-knee orthoses.

All patients with an L4 level could walk whether the hip was stable or not (n=14). Only a third of those with an L3 level walked and the majority of these had un unstable hip. (n=24)

Hip stability, intelligence quotient and fixed deformity did not influence walking ability.

C

- Unclear if inception cohort

- Retrospective cohort study

- FU incomplete (36%)

Iborra, 1999

Type of study: Cross-sectional

Setting: Hospital based

Country: Spain

Source of funding: not reported

Inclusion criteria: MMC patients who were

treated and followed-up in a multidisciplinary unit from 1967- 1995.

Exclusion criteria:

Death (n-17)

Patients for whom control had been

interrupted during the previous 3 years (n=55)

N= 322

Mean age ± SD:
15.9 yr (± 8)

Sex: 55% M /45 % F

Other important characteristics:

Prognostic factor(s) and method of measurement:

Functional neurologic level: the lowest level muscle capable of an antigravitory force of grade 3 or more on the MRC scale.

Patients were grouped according to functional

levels: thoracic (T6 - T12), high lumbar (L1 -L2), mid-

lumbar (L3), low lumbar (L4 - L5) and sacral (S1 - S5).

When relating the different variables to the neurologic

level, the most highly affected of the two sides was chosen.

BMI (Quelet, Roche)

Wechsler Intelligence Scale

Spine deformities (X-rays, criteria Scoliosis,Research Society)

Shunts

Brain lesions

Hip flexion contracture

Sub- or dislocated hips

Endpoint of follow-up: n.a.

For how many participants were no complete outcome data available?

N (%): n.a.

Reasons for incomplete outcome data described? n.a.

Outcome measures and effect size

Initiation of ambulation / age (mo)

Ambulatory status:

Community ambulation: 610 m

Household 30 m

Non-functional < 3 m

Non-ambulators

C

- Unclear if inception cohort

- Retrospective study

- No analysis of prognostic factors related to start of walking

Sival, 1997

Type of study: prospective

Setting: hospital based

Country: The Netherlands

Source of funding: not reported

Inclusion criteria:13 pregnancies complicated by spina bifida aperta

Exclusion criteria:

N=13

Mean age ± SD: prenatal: 18-39 wks

Sex: % M / % F

Not reported

Other important characteristics:

Describe prognostic factor(s) and method of measurement:

Quality of prenatal leg movements and general movements (ultrasound)

Spinal level of morphological defect.

Endpoint of follow-up: 18 mo.

For how many participants were no complete outcome data available?

N (%): 58%

Reasons for incomplete outcome data described?

Yes (death, n=7, n=1 ?)

Outcome measures and effect size (include 95% CI and p-value if available):

Walking y/n

Postnatally, the level of the spinal defect was assessed by means of radiological investigation of the vertebral column

Postnatal sensory function

the level of sensory function was determined by the lowest dermatome at which a pinprick elicited clearly

emotional responses

Postnatal sensory function could be tested in 11 out of 13 cases. Relating the lowest level of sensory function (at day 1) to the prenatally assessed defect (by means of ultrasound) revealed statistical significance (r = 0.76; P = 0.005). In 6 out of 11 cases neonatal death prevented a comparison of sensory function and final neurological outcome. In the remaining 5 neonates sensory function at day 1 proved to be a good indicator for the final neurological dysfunction at 18 months.

A favourable final motor outcome (i.e. walking) was found in neonates (case 9 and 13) with the lowest spinal sensory function. Neonatal leg movements were not related to the spinal defect (r = 0.76; P = 0.005) and to the final motor outcome.

Postnatal motor behaviour

Postnatally, video recordings were made of the infants’ spontaneous motor behaviour and scored off-line for qualitative aspects. When possible, videorecordings were made during the first and second postnatal day, and further on a weekly basis during admission.

Neurological condition assessed repeatedly by an independent child neurologist, not involved in the study.

C

- Small study size (n=13)

Swank, 1992

Type of study: retrospective cohort

(data collection prospective)

Setting: not reported

(same as Swank, 1996) hospital based

Country: USA

Source of funding: not reported

Inclusion criteria: all patients with MMC who presented to our institution within the first 24 hours of life since late 1975

Exclusion criteria:

N=206

Mean age ± SD: 5.12 yr (1 mo – 13 yr)

Sex: % M / % F

Other important characteristics:

Prognostic factor(s) and method of measurement:

Functional motor level:

1)Thoracic level/high lumbar level – no gluteus medius function – (T12-L2); 2)Lumbar level - strong quadriceps and medial hamstrings, no gastrocsoleus or gluteus power (L3-L5); sacral level – quadriceps, medial hamstrings, gastrocsoleus and some gluteus function (S1-3) - Motor level (manual muscle testing) before closure, discharge, 6 mo., yearly

Sitting balance (clinical assessment)

Number and types of deformities (clinic charts, clinical and/or radiographic abnormality)

Orthopaedic procedures (number and indications) (operative reports)

Endpoint of follow-up: 5.12 yr (1 mo -13 yr)

For how many participants were no complete outcome data available?

N (%): 80 (42% thoracic level, 18 % lumbar level and 40% sacral)

Reasons for incomplete outcome data described? Not reported

Outcome measures and effect size

Walking ability (Hoffer, 1973)

Average age of first ambulation was 27.9 mo. T-patients (n-71): 35.1 mo; L-patients (n=73) : 27.7 mo, S-patients (n=62): 22.9 mo. (p=n.s. due to variability and small numbers).

Patients > 48 mo.

24% of T-patients (n=29) were community ambulators, 41% were household ambulators and 35% non ambulatory. L-patients (n=36) 92% c.a., 8% h.a.. S-patients (n=20): all c.a. except one with severe balance deficit.

 None of the patients with moderate or severe balance deficits were c.a. (p=n.s.)

Thoracic, lumbar and sacral level patients had an average of 4.9, 6.3 and 4.0 procedures each.

B

- Retrospective cohort study

Swank, 1994

Type of study: prospective cohort, prediction model

Setting: hospital based

Country: USA

Source of funding: not reported

Inclusion criteria: all children referred to Children’s Memorial Hospital in Chicago, treated with early neurosurgical closure without selective criteria and aggressive correction of their orthopaedic deformities between 1975 and 1985.

Exclusion criteria:

N=206 from an initial cohort of 286 children. 163 patients were followed since 1985, 40 patients were seen with 2 years of the study’s completion. 3 were LTFU

Mean age ± SD: 115 mo (±42)

Sex: % M / % F

Not reported

Other important characteristics:

Prognostic factor(s) and method of measurement:

Motor level, defined on the basis if manual muscle testing by a registered physical therapist.

Thoracic > L3, without functional quadriceps

Lumbar level, strong quadriceps and medial hamstrings, no gastrocsoleus (L3-L5)

Sacral level < S1 and at least some gastrocsoleus function.

Sitting balance deficit. (none, mild, moderate or severe)

Number and type of surgeries.

Presence and number of shunt revisions.

Presence of tethered cord.

Endpoint of follow-up: ?

For how many participants were no complete outcome data available?

N (%): 3, 43 – 22.3%

Reasons for incomplete outcome data described? No

Outcome measures and effect size

Ambulatory status (modified from Hoffer, 1973 – no exercise walking)

Dichotomized: household ambulation and non-ambulation > non-ambulators / community ambulation

The two variables that were statistically correlated with ambulatory status were motor level at birth (p<0.05) and sitting balance (p<0.05)

B

- Prospective inception cohort

- Unclear if follow-up was sufficiently long

- Reasons for missing data not reported

- Adequate control for confounding

Verhoef, 2006

Type of study:

retrospectifive cohort

Setting: Hospital-based

Country: The Netherlands

Source of funding:

Not reported

Inclusion criteria: Young adults with different types of spina bifida (aperta and occulta; ICD); between 16 and 25 years and living in the Netherlands

Exclusion criteria: patients with spina bifida occulta without any neurological loss, non-Dutch-speaking patients,

or patients with comorbidity independently causing more

physical and/or cognitive problems than the spina bifida itself.

Patients were recruited from 11 of the 12 multidisciplinary

spina bifida teams in the Netherlands. The Dutch Association for Patients with Spina Bifida also invited members to participate.

In addition, rehabilitation centres, special housing facilities,

and special schools were approached to find potential

participants.

N= 168

Data for the present study on functional independence

were complete for 165 patients.

Mean age ± SD: 20 yr 9 months (range 16 -25 yr ± 2.9)

Sex: 42 % M /58 % F

84% had spina bifida aperta

71% of patients had hydrocephalus.

44% of patients had a level of lesion of L2 and above, 40%

L3 to L5, and 16% S1 and below.

functional independence among young adults with spina bifida and its relationships with pathological characteristics known from birth (hydrocephalus and level of lesion).

 Functional independence was assessed on six domains (self-care, sphincter control, transfers, locomotion, communication, and social cognition) using the Functional Independence Measure (FIM).

Endpoint of follow-up: ?

For how many participants were no complete outcome data available?

N =11

Reasons for incomplete outcome data described?

because no neurological loss was determined; two were excluded because of comorbidity independently

inducing serious physical and/or cognitive disorders

(one had a serious heart disease and one a chromosome

disorder).

Patients without hydrocephalus were independent for all FIM domains except sphincter control, as were patients with hydrocephalus with a lesion level below L2.

 Most patients with hydrocephalus and a lesion at L2 or above were dependent as regards sphincter control (98%), locomotion (79%), and self-care (54%), and quite a few needed support in transfers (38%), social cognition (29%), and communication (15%).

C

- retrospective cohort

- Unclear if follow-up was sufficiently long

- small amounts

- Adequate control for confounding

Williams, 1999

Type of study:

Retrospective cohort

Setting: Hospital-based

Country: Australia

Source of funding:

Not reported

Inclusion criteria: Patients with SB who were regularly assessed at the Royal’s Children’s Hopsital Melbourne between 1978-1993.

Diagnoses included

MMC, lipomyelomeningocele, or intraspinal lipoma.

Community ambulation or cessation sustained for more than one evaluation

Exclusion criteria:

Asymmetry significant enough to affect ambulation (n=1). when lesion level could not be determined

because of spasticity. Children under 5 years of age were only

included if recorded as cooperative when manual muscle

charting was performed (McDonald et al. 1986). Exceptions

were made for 3 others who were included, although not

recorded as cooperative, as they clearly had a lesion at the sacral level, were ambulating by 3 years of age, and met the

other criteria.

N=173

Mean age ± SD: 3-9 yr (3-22)

Sex: % M / % F

not reported

Prognostic factor(s) and method of measurement:

Neurosegmental level (International Myeolodysplasia Study Criteria Manual), and modified from Sharard (1964)

Endpoint of follow-up:

9 yr = mean age of latest evaluation

At least 3 regular evaluations at which a diagnosis, age in months, cooperative level, motor (lesion) level, and usual

means of locomotion were recorded.

For how many participants were no complete outcome data available?

N (%): n.a.

Reasons for incomplete outcome data described?

n.a.

Outcome measures and effect size

Community ambulation (defined as complete ambulatory function with or without aids or braces, wheelchair may be used for long distances, Broughton, 1993)

Thirty-five children had a lesion at the thoracic level; seven

of whom walked at a mean age of 4 years 6 months (20%). .

Ten children had a lesion at the high-lumbar level (L1/2) and five

walked at a mean age of 5 years 2 months. (50%)

Fifteen had a lesion at the mid-lumbar level (L3) and nine walked at a mean age of 5 years. (60%)

Forty-five children had a lesion at the low-lumbar level (L4/5) and 38 walked at a mean age of 3 years 10 months. (84%)

Sixty-eight children had a lesion at the sacral level; all of whom all

walked at an average age of 2 years 2 months. (100%)

C

- Unclear if inception cohort

- Retrospective study

- No analysis of prognostic factors related to start of walking

Study reference

Study characteristics

Patient characteristics

Prognostic factor(s)

Follow-up

Outcome measures and effect size

Study quality

Research question: What factors are associated with maintenance or loss of ambulation in ambulatory children and adolescents with spina bifida?

Asher, 1983

Type of study:

Cohort study (unclear whether retrospective or prospective)

Setting:

Hospital

Country:

USA

Source of funding:

Not stated

Inclusion criteria:

patients with spina bifida cystica treated and followed at University of Kansas Medical Center

Exclusion criteria:

Severe concurrent spastic tetraplegia

N=98

Mean age and range:

14 years and 4 months (6-32 years)

Sex: 51% M / 49% F

Other important characteristics:

Thoracic (T10-12):

28 (29%)

High lumbar (L1-2):

12 (12%)

Mid lumbar (L3):

20 (20%)

Low lumbar (L4-5):

26 (27%)

Sacral:

12 (12%)

Describe prognostic factor(s) and method of measurement:

- Neurological level

- Age at time of study

- Age child was first fitted with braces and stood

- Age at closure of sac

- Hydrocephalus

- Seizures or spasticity

- Spine deformity

- Pelvic obliquity

- Hip deformity

- Knee-foot-ankle deformity

- Fractures

- Kidney status

- Weight

- IQ

- Psychological dependence

Endpoint of follow-up:

Unclear

For how many participants were no complete outcome data available?

N (%):

0

Reasons for incomplete outcome data described?

Not applicable

Outcome measures and effect size (include 95%CI and p-value if available):

Most significant single factor affecting walking ability was level of neural involvement (p < 0.001)

Significant variables associated with walking ability by neural level (Pearson linear correlation coefficient):

T:

- Age at time of study (p < 0.001)

- knee-foot-ankle deformities (p < 0.005)

L1-2:

- Weight (p < 0.01)

L3:

- hip deformity (p < 0.001)

L4:

- Age when first braced (p < 0.001)

- spinal deformity (p < 0.005)

- pelvic obliquity (p < 0.001)

- hip deformity (p < 0.001)

- knee-foot-ankle deformities (p < 0.005)

L5 or sacral:

all (but 1) community ambulators

Transitions in ambulatory status (only descriptive, no statistical analysis):

T:

11/21 patients had transitions:

2 upward, 17 downward (motivation of patient)

L1-2:

6/12 patients had transitions: 4 upward (improved orthotic management), 13 downward (lack of motivation, skin problems, obesity)

L3:

16/20 patients had transitions: 3 upward (improved orthotic management), 19 downward (lack of motivation, skin problems, obesity, musculoskeletal deformity)

L4:

3/17 patients had transitions: 3 upward (correction of musculoskeletal deformity), 2 downward (increasing musculoskeletal deformity)

L5 or sacral:

No transitions

Upward transitions most commonly occurred between 5 and 10 years of age. There was no common age for downward transitions.

C

- Unclear whether prospective or retrospective cohort

- Unclear if inception cohort

- Endpoint of follow-up not reported

- No control for confounding (multivariate analysis)

- Only narrative analysis of factors possibly associated with changes in ambulation

Bartonek,

1999

Type of study:

Prospective cohort study

Setting:

Hospital

Country:

Sweden

Source of funding:

grants from Norrbacka Eugenia Foundation and Foundation "Konung Oscar II och Drottning Sophias Guldbröllopsminne"

Inclusion criteria:

ambulating patients with myelomeningocele treated between 1952 and 1983 at the Karolinska Hospital in Stockholm and Örebro Medical Center Hospital

Exclusion criteria:

Not reported

N=100, 60 available for follow-up

Median age and range:

22 years (12-54)

Sex: 43% M / 57% F

Other important characteristics:

41 community, 12 household, 7 nonfunctional ambulators

Describe prognostic factor(s) and method of measurement:

- Neurological level

- Spasticity lower limbs

- Hip-flexion contracture ≥ 20°

- Knee-flexion contracture ≥ 20°

- Scoliosis ≥ 20°

- Major medical events (eg stroke, lower limb edema, recurrent skin infections, frequent epileptic seizures, multiple shunt revisions, invasive surgical prodedures)

- Pressure sores

 - Ischial tuberosity

 - Foot

- Musculoskeletal pain

- Use of orthoses

Endpoint of follow-up:

12 years

For how many participants were no complete outcome data available?

N (%):

N=40 (40%)

Reasons for incomplete outcome data described?

N=14 untraceable

N=1 relocated

N=25 refused to participate

Outcome measures and effect size (include 95%CI and p-value if available):

No statistical analysis, only comparison of factors between patients with unchanged ambulation level (N=41) vs. patients with deteriorated ambulation level (N=19)

Neurological level:

 17% vs. 47%

Spasticity lower limbs:

20% vs. 26%

Hip-flexion contracture ≥ 20°:

15% vs. 79%

Knee-flexion contracture ≥ 20°:

5% vs. 89%

Scoliosis ≥ 20°:

24% vs. 53%

Major medical events:

15% vs. 89%

Pressure sores

- Ischial tuberosity:

2% vs. 21%

- Foot:

12% vs. 0%

Musculoskeletal pain:

49% vs. 63%

Use of orthoses:

20% vs. 58%

C

 - Unclear if inception cohort

- 40% loss to follow-up

- No statistical analysis (and no control for confounding)

- No analysis of factors associated with changes in ambulation

Buffart, 2008

Type of study:

Cross-sectional study

Setting:

University hospitals and rehabilitation centres

Country:

Netherlands

Source of funding:

Johanna Children’s Fund

Inclusion criteria:

- Adolescents and young adults with MMC, aged between 16 and 30 years from 4 university hospitals

and 5 rehabilitation centres in the adjacent region

Exclusion criteria:

- inability to understand the measurements performed in this study (judged by the treating physician or a family member)

- complete dependence on a powered wheelchair

- presence of disorders other than MMC that affect daily physical activity

- contra-indication for a maximal exercise test

N=51

Mean age ± SD:

21.1 ± 4.5 years

Sex: 51% M / 49% F

Other important characteristics:

15 community, 8 household, 28 (non) functional ambulators

Describe prognostic factor(s) and method of measurement:

- daily physical activity: AM

- aerobic fitness: maximum oxygen uptake during progressive maximal aerobic test on arm or cycle ergometer

- body composition and body fat: sum of 4 skin fold measurements, BMI

Endpoint of follow-up:

Not applicable

For how many participants were no complete outcome data available?

N (%):

0

Reasons for incomplete outcome data described?

Not applicable

Outcome measures and effect size (include 95%CI and p-value if available):

Multiple regression analyses performed. β = standardized

regression coefficient

Ambulatory status was related to daily physical activity β = 0.541, p<0.001), aerobic fitness (β = 0.397, p<0.001) and body fat (β = –0.243, p=0.03).

Physical activity was related to aerobic fitness in non-ambulatory persons with MMC (β = 0.398, p=0.04), but not in ambulatory persons.

Body fat was not related

to physical activity, or to aerobic fitness in both ambulatory

and non-ambulatory persons with MMC.

D

- Cross-sectional study design (no causal relationships)

Fraser, 1992

Type of study: retrospective cohort

Setting: hospital based

Country: South Africa

Source of funding: not reported

Inclusion criteria: Patients presenting with mid-lumbar neurological MMC during 1970-1986.

Exclusion criteria: ≤4 yr

N=55

Mean age ± SD: not reported

Sex: % M / % F not reported

All ambulant patients used bilateral elbow crutches and below-knee orthoses.

Prognostic factor(s) and method of measurement:

Hip stability

Fixed flexion deformities ≥10◦, deformities of spine, ankle or foot.

Spasticity in lower limbs

Intelligence

Neurological lesion level:

L3 quadriceps power of MRC ≥ grade 3 medial hamstrings ≤ grade 3 (n=24)

 L4 :quadriceps power ≤ MRC grade 3, medial hamstrings ≤ grade 3. (n=14)

Radiography of hips, dislocated or subluxated hips were considered to be unstable.

Orthotic use

Surgical procedures

Endpoint of follow-up: 10.2 yr (4-19)

For how many participants were no complete outcome data available?

N (%): 20 (36%)

Reasons for incomplete outcome data described?

Yes (n=7), no (n=13) – 24%

Outcome measures and effect size

Neurological most significant determinant of walking ability:

All patients with L4 level could walk, but only one-third of those with L3 lesion could do so.

Hip stability, intelligence quotient and fixed deformity did not influence walking ability.

C

- 36% loss to follow-up

- No statistical analysis (and no control for confounding)

- No analysis of factors associated with changes in ambulation

McDonald, 1991

Type of study: retrospective cohort study

Setting: hospital based

Country: USA

Source of funding: spina bifida association of America, March of Dimes-Birth Defects Foundation, Student Research and Computer Funds

Inclusion criteria:

Patients with myelomeningocele and at least 3 evaluations after the age of 5 years

Exclusion criteria:

Not reported

N=308 (from clinic population of 981 patients seen between 1957 and 1985)

Mean age at last examination (median): 174 months=14.5 years (171 months)

Sex: not reported

Neurological level:

Thoracic: 21.5%

Upper lumbar: 9.8%

Mid-lumbar: 17.1%

Lower lumbar: 25.9%

Sacral: 25.7%

Prognostic factor(s) and method of measurement:

Lower-extremity muscle strength tested with standard manual muscle testing (0-5 or intermediate according to guidelines used by International Myelodysplasia Study Group.

Muscle strength was recorded as the median of all examinations after age 60 months and before any surgical procedures which potentially weaken muscle strength.

Ambulatory status:

nonambulatory, partial ambulator, community ambulator with aids and/or braces but no wheelchair) or community ambulator without aids or braces.

Endpoint of follow-up: mean age 174 months

For how many participants were no complete outcome data available?

N (%): 17/308=5.5%

Reasons for incomplete outcome data described?

Progressive loss of muscle strength (≥3 grades in more than 1 muscle) because of Arnold-Chiari malformation or because muscle strengths were unreliable because of limited joint movement or spasticity in all lower-extremity muscles

Outcome measures and effect size

Stepwise multiple regression analysis:

Independent variables that contributed statistically significant (p< 0.05) to ambulatory status:

• iliopsoas
• gluteus maximus
• gluteus medius
• quadriceps
• anterior tibialis

Muscle strength > 3 positively correlated with ambulation.

Changes in mobility were not age-related (no significant differences in changes across all ages, and they occurred in both directions).

C

- Only limited analysis of factors possibly associated with changes in ambulation over time (limitations: few patients had serial examinations across all age ranges of interest, cross-sectional analyses are inappropriate because of historical differences in the care given to the various age cohorts, no analysis of potential other causal effects)

Oakeshott, 2003

Type of study:

Prospective cohort study

Setting:

Hospital

Country:

England

Source of funding:

Association for Spina Bifida and Hydrocephalus (ASBAH)

Inclusion criteria:

Newborn babies with open spina bifida referred to the Regional Neurosurgical Unit at Addenbrooke’s Hospital, Cambridge between 1963 and 1971

Exclusion criteria:

Not reported

N=117, 57 survivors at follow-up

Mean age (range):

30 years (26-33) (N=57)

Sex: 44% M / 56% F (N=57)

Other important characteristics:

Sensory level at infancy:

13 above T11, 16 T11-L3, 8 L4, 6 L5-S2, 11 no sensory loss, 3 asymmetrical loss

Describe prognostic factor(s) and method of measurement:

Sensory level

Endpoint of follow-up:

Mean of 30 years

For how many participants were no complete outcome data available?

N (%):

60/117 (51%)

Reasons for incomplete outcome data described?

N=60 deaths

Outcome measures and effect size (include 95%CI and p-value if available):

Community walkers at 9 years (n=32) vs. community walkers at 30 years (n=17):

Above T11: 0 vs 0

T11-L3: 5 vs. 0

L4: 8 vs. 1

L5-S2: 6 vs. 5

No sensory loss: 11 vs. 10

Asymmetrical loss: 2 vs. 1

Survivors with lower sensory levels are more likely to be walkers (X²

 for trend P<0.0001 for both age nine years and 30 years).

Community walkers at 9 years: 32/57 (56%)

Community walkers at 30 years: 17/57 (30%)

Community walker: being albe to walk > 50 metres with or without aids

C

- No control for confounding (multivariate analysis)

- No analysis of factors associated with changes in ambulation

Samuelsson, 1988

Type of study:

Prospective cohort study

Setting:

hospital

Country:

Sweden

Source of funding:

Orebro county council and king oscar’s and queen sophia’s golden wedding memorial fund

Inclusion criteria:

Whole population of patients with MMC treated at Karolinska Hospital and Orebro Medical Center between 1952 and 1983

Exclusion criteria:

Not reported

N=196 whole population, 163 included in analysis

Mean age ± SD:

14 years and 10 months (2-40)

Sex: 85 M / 78 F

Describe prognostic factor(s) and method of measurement:

Neurological level

Scoliosis

Pelvic obliquity

Hip disclocation

Flexion contracture at hip and knee

Endpoint of follow-up:

Not reported

For how many participants were no complete outcome data available?

N (%):

36/196 (18%)

Reasons for incomplete outcome data described?

N=11 death

N=17 relocated

N= 5 untraceable

N=3 unknown

Outcome measures and effect size (include 95%CI and p-value if available):

Amount of scoliosis between ambulators and nonambulators at each neurological level was significantly different (p < 0.05)

C

- No control for confounding (multivariate analysis)

- No analysis of factors associated with changes in ambulation&