Early detection of dysphagia and dysarthria in children with neuromuscular disorders: Diagnostic accuracy of a Screeninglist for Physicians

Abstract

PURPOSE:

Regardless of age or disease stage, children with neuromuscular disorders (NMD) are at risk of developing dysphagia and/or dysarthria. It is important to screen these children regularly in order to detect and treat problems as soon as possible. To date, there are no standardized tools for screening for dysphagia and dysarthria in children with NMD (pNMD). Thus, children are not always referred for assessment by a speech language therapist (SLT). A new screening instrument for dysphagia and dysarthria has been developed, the Screeninglist Physician of the Diagnostic list for Dysphagia and Dysarthria in pediatric NMD (DDD-pNMD). The diagnostic accuracy was estimated in this study.

METHODS:

Sensitivity and specificity were assessed in 131 children aged 2.0–18.0 years by comparing the outcome of the Screeninglist Physician with the diagnosis of dysphagia and/or dysarthria established by an SLT.

RESULTS:

The sensitivity of the Screeninglist Physician was 88% and its specificity was 63%. The AUC was 0.83. The prevalence of dysphagia and/or dysarthria was 53%.

CONCLUSION

: The Screeninglist Physician of the DDD-pNMD is the first valid screening tool for physicians to identify children with NMD with possible dysphagia and/or dysarthria, thereby enabling timely referral to an SLT.

Keywords

Dysphagia dysarthria neuromuscular disorders children screening sensitivity diagnostic accuracy

1. Introduction

Neuromuscular disorders (NMDs) are a broad group of hereditary or acquired muscle disorders of variable severity. There are about 600 different muscle diseases which are often classified based on their anatomical origin: in the ventral horn of the spinal cord, the peripheral nerves, the neuromuscular junction, or the muscle fibers. There are congenital, stationary and slowly or rapidly progressive forms of NMDs [1]. Modern diagnostic procedures and improved medical management and rehabilitation therapy have increased the life expectancy of children with NMD [2]. However, swallowing and speech difficulties remain common problems in these individuals [3].

Normal swallowing includes oral, pharyngeal and esophageal phases [4]. Problems in one or more of these phases are called dysphagia, which may occur in all pediatric NMD (pNMD) [3] and are usually caused by weakness of the orofacial, pharyngeal, and thoracic muscles [5]. Its presence and severity largely depend on the muscle disease concerned [3, 6]. Children with NMD can have problems with chewing and need to swallow frequently in order to achieve complete emptying of the oral cavity and pharyngeal area, often resulting in long mealtimes [2, 5, 9]. These problems are caused by a decreased bite force and weakness of the masticatory muscles. Limited jaw opening affects chewing and oral hygiene [7], and many children retain food in the pharynx after swallowing [8]. The consequences of dysphagia can be substantial: failure to thrive, malnutrition, choking, recurrent respiratory infections, and aspiration pneumonia [10, 11, 12].

Children with NMD may also experience speech problems. In speech, many orofacial and thoracic muscles are involved, all contributing to strength, speed, range, steadiness, tone, or accuracy of movements necessary for intelligible and natural speech. Motor speech disorders called dysarthria are caused by disease of the peripheral and/or central nervous system which can affect various aspects of speech such as respiration, phonation, articulation, nasal resonance, and/or prosody [13, 14]. The type of dysarthria depends on the location of the injury in the nervous system [15]. Common features of dysarthria in pNMD include weak articulation, hypernasality, weak phonation, flattened prosody, and reduced speech volume [16, 17, 18, 19, 20].

Dysphagia often causes malnutrition, choking, or aspiration pneumonia [6, 21] and dysarthria influences intelligibility [22]; both hamper a child’s ability to participate. For this reason, it is important to monitor dysphagia and dysarthria throughout development. In the Netherlands, all children diagnosed with pNMD are regularly assessed by a multidisciplinary rehabilitation team under the direction of a rehabilitation physician or child neurologist. However, a speech language therapist (SLT) is not always part of this team in rehabilitation centers or hospitals. Therefore, possibly not all children are annually assessed by an SLT. Physicians are responsible for indicating whether referral to an SLT is necessary. For this reason, a validated, standardized tool for screening of dysphagia and dysarthria in pNMD that can be used by the physician was needed to create a better opportunity for referral [21, 23]. There is a wide range of problems due to muscle weakness that can occur in pNMD in which symptoms of dysphagia and dysarthria can be difficult to spot by physicians. Therefore, a Dutch working group of experienced SLTs involved in the care of children with NMD developed the Screeninglist Physician as part of a larger diagnostic instrument, the Diagnostic list of Dysphagia and Dysarthria in NMD (DDD-NMD) [24]. The intended use of the list was to screen for signs of dysphagia and dysarthria. If a child fails the screening, referral to an SLT for an extensive assessment should follow.

The aim of this study was to estimate the diagnostic accuracy of the Screeninglist Physician of the DDD-NMD compared with a clinical reference standard (SLT assessment and diagnosis) in children aged 2–18 years with an NMD. For this purpose, sensitivity, specificity, predictive values, area under the ROC-curve and the cut-off point were measured.

2. Method

A cross-sectional prospective study was carried out involving 131 children (aged 2.0–18.0 years) with an NMD who visited the pediatric multidisciplinary outpatient clinic of the Muscle Center of the Radboud University Medical Centre in Nijmegen or Rehabilitation Centre Klimmendaal in Arnhem (both in The Netherlands) between May 2015 and October 2016. SLT’s are a standard part of the multidisciplinary pNMD-teams of both centers. All consecutively admitted patients were eligible for inclusion if: (i) they had been diagnosed with an NMD; and (ii) were between 2;0 and 18;0 years of age.

2.1 Screeninglist Physician of the DDD-pNMD

The Screeninglist Physician is a short and quick screening, consisting of nine broad yes/no-questions (to be answered by parents/caregivers or by the young adult patient) (see Fig. 1): seven questions on feeding and swallowing and two questions on speech and communication. The broadly formulated questions are related to frequently reported complaints in pNMD [3].

Figure 1.

Questions of the Screeninglist Physician (DDD-pNMD) (translated from Dutch).

The physician administered the Screeninglist Physician together with caregivers or with young adults when visiting the outpatient clinic. The test positive cut-off was set at $\geqslant$ 1. The SLT had no knowledge of the completed Screeninglist Physician score.

2.2 Reference standard of diagnosing dysphagia and dysarthria

All children, regardless of the outcome on the Screeninglist Physician, were seen by the SLT for a full assessment. The SLT’s were blinded for the outcome of the Screeninglist Physician. Diagnosis of dysphagia and/or dysarthria by the SLT was used as reference standard. This diagnosis was based on the SLT’s clinical evaluation, following national and international guidelines [25, 26]. To structure the clinical evaluation, the standardized diagnostic part of the DDD-pNMD was completed by an SLT for all children as best practice in the absence of other validated diagnostic tools [21]. The diagnostic part of the DDD-pNMD consists of an extensive speech language assessment covering case history and items of the themes: posture, oral and facial anatomy, oral structure and oral motor skills, saliva control, eating, drinking, chewing, swallowing, intelligibility of speech, and communication. All separate items within these themes are scored on a 4-point Likert scale based on the Radboud Dysarthria Assessment, ranging from 0 to 3 corresponding with no, mild, clear or severe abnormality [27]. A qualitative judgement has to be given in case the score was $\geqslant$ 1. Scoring directions are given in the user manual. The DDD-pNMD is developed in a focus group of experienced SLT’s through consensus [24]. The SLT made an overall conclusion in terms of dysarthria and/or dysphagia: yes/no. If dysphagia and/or dysarthria occurred, an intervention plan was always offered (including instrumental evaluation and treatment).

Table 1
Demographics and clinical characteristics of the participants ( $N=$ 131)

$N$	131
Age range in years; months (mean)	2.1–17.9 (10.5)
Gender, $n$ (%)
Male	94 (72.0)
Female	37 (28.0)
NMD, $n$ (%)
Anterior horn cell	6 (4.6)
Nerves fiber	9 (6.9)
Neuromuscular junction	1 (0.8)
Muscle, divided into:	115 (87.8)
Muscular dystrophy	60 (45.8)
Myotonic dystrophy	13 (9.9)
Congenital myopathy	32 (24.4)
Metabolic myopathy	10 (7.6)

Data were collected from the Screeninglist Physician (positive/negative) and from the conclusion of the SLT assessment (dysarthria: yes/no, dysphagia: yes/no). In total, four physicians and four SLTs (Radboud University Medical Center and Rehabilitation Centre Klimmendaal) were involved in the study. All had more than 5 years clinical experience of NMD in children.

Where appropriate, the standards for reporting diagnostic accuracy studies (STARD) were followed. STARD is an internationally acknowleged guideline developed to help improve the transparency and completeness of the reporting of diagnostic accuracy studies [28]. This study was approved by the Committee on Research Involving Human Subjects of Arnhem and Nijmegen (no 2015–1690). Informed consent was obtained from the parents and children, or only from the parents for children younger than 12 years.

Table 2

Cross tabulation of the indextest results (Screeninglist Physician) by the results of the reference standard (SLT assessment)

		Dysphagia/dysarthria diagnostics by SLT (reference standard)		Total
		Present	Absent
Screeninglist physician	Positive	61 ${}^{\rm a}$	23 ${}^{\rm b}$	84	PV+ $=$ a/(a+b) $=$ 73% (95% C.I. 65–79)
(indextest)	Negative	8 ${}^{\rm c}$	39 ${}^{\rm d}$	47	PV- $=$ d/(c+d) $=$ 83% (95% C.I. 71–91)
Total		69	62	131
		Se $=$ a/a+c $=$ 88% (95% C.I. 78–95)	Sp $=$ d/b+d $=$ 63% (95% C.I. 50–75)		Prev $=$ (a+c)/(a+b+c+d) $=$ 53% (95% CI 44–61)

a $=$ true positive, b $=$ false positive, c $=$ false negative, d $=$ true negative, Se $=$ sensitivity, Sp $=$ specificity, PV+ $=$ positive predictive value, PV- $=$ negative predictive value, Prev $=$ Prevalence.

2.3 Statistical analysis

Descriptive statistics were used to document the general characteristics of the children. Diagnostic test accuracy was defined by sensitivity, specificity, and predictive values with 95% confidence intervals (CIs). Sensitivity, specificity and positive predictable value have been calculated for 3 age groups. A receiver operating characteristic curve (ROC) was used to calculate the area under the ROC. We used the theoretical a priori cut-off score of $\geqslant$ 1 on the Screeninglist Physician. There were no missing data. A cross-tabulation of the index assessment by the reference assessment was produced. Data were analyzed with SPSS 23.0 for Windows (SPSS Inc., Chicago, IL, USA).

3. Results

In total, 137 Dutch children with NMD participated in the study. After the exclusion of non-eligible children, the data of 131 children were analyzed (Table 1, Fig. 2). The SLT assessment was performed within 1 week of the Screeninglist Physician test.

Figure 2.

Flow diagram: flow of participants.

With a theoretical cut-off score of $\geqslant$ 1, the Screeninglist Physician detected symptoms of dysphagia and/or dysarthria in 84 of the 131 children (64%). The SLT diagnosed dysphagia and/or dysarthria in 69 of the 131 children (53%). Eight children (6%, 6 with dysphagia, 2 with dysarthria) were not identified by the Screeninglist Physician. In the whole sample, the sensitivity of the Screeninglist was 88% (95% C.I. 78–95) and its specificity was 63% (95% C.I. 50–75) (see Table 2). The proportion of children identified correctly by the screening instrument as not having dysarthria and/or dysphagia was 83% (95% C.I. 71–91) (negative predictive value). The positive predictive value was 73% (95% C.I. 65–79). The area under the ROC curve (AUC) was 0.83 (95% C.I. 76–90) (see Fig. 3). The prevalence of dysphagia and/or dysarthria was 53% (95% C.I. 44–61); a cross tabulation of dysphagia and dysarthria diagnosis is shown in Table 3. The data of 3 age groups on sensitivity, specificity and the percentage of unnecessarily referred children (1-PPV) is shown in Table 4. In the whole group ( $n=$ 131), 27% of the children were unnecessarily referred, with 33% in the group of children $<$ 7 years ( $n=$ 29), 32% in the 7–12 year group ( $n=$ 55) and 14% in the group $\geqslant$ 12 ( $n=$ 47).

Table 3

Crosstabulation of dysphagia and dysarthria diagnosis by the SLT

	Dysphagia
	No	Yes	Total
Dysarthria
No	62	44	106
Yes	5	20	25
Total	67	64	131

Table 4

Sensitivity, specificity and the percentage unnecessarily referred children (1-PPV) in the total group and 3 age groups

Age range	N	Sensitivity	Specificity	1-PPV
2.1–17.9	131	88	63	27
2.1–6.11	29	94	33	33
7.0–11.11	55	87	52	32
12.0–17.9	47	86	88	14

Figure 3.

Receiver operating characteristic (ROC) curve and area under curve (AUC) for the Screeninglist Physician versus diagnosis by a speech language therapist ( $n=$ 131). AUC $=$ 0.83, 95% CI [0.76–0.90]. Black circle indicates cut-off score of 1 on the Screeninglist Physician (sensitivity 88%, specificity 63%).

4. Discussion

The Screeninglist Physician of the DDD-pNMD is a short and simple screeninglist for dysphagia and dysarthria and was developed by a group of experienced SLTs in the field of pNMD.

Based on the available literature, this is the first screening that is evaluated for diagnostic accuracy for this patient population. Our findings show that the Screeninglist Physician, with an AUC of 0.83 (95% CI 0.76–0.90), has a very good diagnostic accuracy [29], indicating a very good ability to distinguish between children with or without signs of dysphagia and/or dysarthria, diagnosed by an SLT. The physicians found the list clear and easy to use.

The desired sensitivity and specificity characteristics of a screening instrument depend on the purpose of screening. For the Screeninglist Physician, we aimed to focus on a high sensitivity ( $>$ 80%) as a guideline, because it is important not to miss children with genuine problems. Specificity, however, is acceptable at $>$ 50%, because a consequence of a false-positive result is that the child is referred for a further evaluation by an SLT, which is not very invasive, risky, or costly. The Screeninglist Physician’s sensitivity of 88% and specificity of 63% imply that the instrument is sensitive enough to detect children with potential dysphagia and/or dysarthria and is specific enough for screening purposes. Six percent ( $n=$ 8) of 131 children with problems were missed by the Screeninglist Physician. Six of these children were having chewing problems. To detect these problems a question could be added to the list related to this topic (‘Does the child require food to be cut because the child cannot adequately chew?’).

Due to the sensitive and the less specific nature of the screening, children might be referred unnecessarily. Comparing 3 age groups ( $<$ 7, 7–12 and $\geqslant$ 12 years), we found that 33% and 32% of the children in the youngest groups were referred unnecessarily. This is slightly higher than the 14% in the oldest group, possibly because the list doesn’t account for the developmental phases in speech of younger children. In the youngest age group, specificity did not meet our criteria of $>$ 50%. That means that the Screeninglist Physician is not reliable enough for screening children younger than 6;11 and that the results must be carefully interpreted. Explanations for the low specificity in the youngest age group might be developmental issues like speech and swallowing tasks taken for dysarthric of dysphagic components. In the oldest 2 groups (7.0–11.11 and 12.0–17.9 years), both sensitivity and specificity were within our criteria.

The items on the Screeninglist Physician are covering a broad spectrum of symptoms of dysphagia. The cut-off score of $\geqslant$ 1 was based on clinical experience and knowledge of the authors, by reasoning that each of these items can lead to a positive diagnosis. The ROC (Fig. 1) showed that this theoretical point was indeed the most accurate cut-off point.

Screening is important for multiple reasons. First, the prevalence of dysphagia and dysarthria is high in children with NMD. For example, in children with SMA II and III it is 44.4% [6], and in children with myotonic dystrophy it is 51.9% [5]. In our study population, the prevalence of dysphagia and dysarthria was 53%. Second, dysphagia can have severe, sometimes life-threatening consequences [21], such as malnutrition, choking, and aspiration pneumonia. Dysarthria might hamper the function and participation of affected children and therefore adversely affect their quality of life. The high prevalence, in combination with the variable onset and consequences of dysphagia and dysarthria, justifies periodic monitoring of these children which would be greatly facilitated by using a short and easy to use screening instrument such as the Screeninglist Physician.

Due to measurements of diagnostic accuracy being sensitive to the design of the study, the STARD checklist was followed to optimize the design and reporting. Nevertheless, there are some limitations that must be taken into consideration.

First is that we had to rely on a best practice reference assessment and diagnosis by experienced SLTs in the absence of a comprehensive proven, valid and reliable instrument for the diagnosis of dysphagia and dysarthria in pNMD [21, 23]. Instrumental assessment such as a videofluoroscopic swallowing study (VFSS) is considered as the gold standard for detection of pharyngeal dysphagia. However, chewing problems or prolonged mealtimes are commonly reported complaints of dysphagia in pNMD [3] causing failure on the Screeninglist Physician. Children showing chewing problems are not referred for instrumental assessment, because these oral phase problems are not captured by VFSS. Second, a VFSS gives a small window of swallowing ability in which the common factor like fatigue is not included. Children with pNMD often cannot maintain ‘best swallow’ during the entire mealtime [21]. To establish a diagnosis of dysphagia and dysarthria, a broader assessment is needed. This is the reason why the SLT assessment was chosen as the best available (best reference standard) method for establishing the presence or absence of dysphagia or dysarthria.

If there were concerns about safe swallowing, or the feeling of food sticking in the throat, children were always referred for instrumental assessment as part of the assessment and intervention plan.

Second, test-retest reliability was not measured because the bias caused by awareness of the parents and possible progression of the disease could influence the results.

5. Conclusion

The Screeninglist Physician of the DDD-pNMD is the first valid tool to screen children and young adults with NMD for possible dysphagia and/or dysarthria, thereby enabling timely referral to an SLT.

Footnotes

Conflict of interest

The authors have no conflict of interest to report.

References

Dubowitz

. Muscle disorders in childhood. 2nd ed. London: WB.Saunders Company Ltd., 2000.

Ramelli

, Aloysius

, King

, Davis

, Muntoni

. Gastrostomy placement in paediatric patients with neuromuscular disorders: indications and outcome. Dev Med Child Neurol. 2007; 49(5): 367-71. doi: 10.1111/j.1469-8749.2007.00367.x.

van den Engel-Hoek

, de Groot

IJM

, de Swart

BJM

, Erasmus

. Feeding and swallowing disorders in pediatric neuromuscular diseases: an overview. J Neuromuscul Dis. 2015; 2(4): 357-69. doi: 10.3233/JND-150122.

Miller

. The neurobiology of swallowing and dysphagia. Dev Disabil Res Rev. 2008; 14: 77-86. doi: 10.1002/ddrr.12.

Sjogreen

, Engvall

, Ekstrom

, Lohmander

, Kiliadiris

, Tulinius

. Orofacial dysfunction in children and adolescents with myotonic dystrophy. Dev Med Child Neurol. 2007; 49(1): 18-22. doi: 10.1017/S0012162207000060.

Chen

, Shih

, Chen

, Kuo

, Jong

. Prevalence and risk factors for feeding and swallowing difficulties in spinal muscular atrophy types II and III. J Pediatr. 2012; 160(3): 447-51.

van Bruggen

, van den Engel-Hoek

, van der Pol

, de Wijer

, de Groot

IJM

, Steenks

. Impaired mandibular function in spinal muscular atrophy type II: need for early recognition. J Child Neurol. 2011; 26(11): 1392-6. doi: 10.1177/0883073811407696.

van den Engel-Hoek

, Erasmus

, Hendriks

JCM

, Geurts

ACH

, Klein

, Pillen

, et al. Oral muscles are progressively affected in Duchenne muscular dystrophy: implications for dysphagia treatment. J Neurol. 2013; 260(5): 1295-303. doi: 10.1007/s00415-012-6793-y.

Tieleman

, Knuijt

, van Vliet

, De Swart

BJM

, Ensink

, van Engelen

BGM

. Dysphagia is present but mild in myotonic dystrophy type 2. Neuromuscul Disord. 2009; 19(3): 196-8. doi: 10.1016/j.nmd.2008.12.002.

10.

Philpot

, Bagnall

, King

, Dubowitz

, Muntoni

. Feeding problems in merosin deficient congenital muscular dystrophy. Arch Dis Child. 1999; 80(6): 542-7.

11.

Prasse

, Kikano

. An overview of pediatric dysphagia. Clin Pediatr (Phila). 2009; 48(3): 247-51. doi: 10.1177/0009922808327323.

12.

Taniguchi

, Moyer

. Assessment of Risk-factors for pneumonia in dysphagic children – significance of videofluoroscopic swallowing evaluation. Dev Med Child Neurol. 1994; 36(6): 495-502. doi: 10.1111/j.1469-8749.1994.tb11879.x.

13.

Pennington,

, Parker,

, Kelly,

, Miller,

. Speech therapy for children with dysarthria acquired before three years of age. Cochrane Database of Syst Rev. 2016; (7). doi: 10.1002/14651858.CD006937.

14.

Duffy,

. Motor speech disorders: Substrates, differential diagnosis, and management. St. Louis, MO: Elsevier. 2013.

15.

Pennington

, Miller

, Robson

, Steen

. Intensive speech and language therapy for older children with cerebral palsy: a systems approach. Dev Med Child Neurol. 2010; 52(4): 337-44. doi: 10.1111/j.1469-8749.2009.03366.x.

16.

Bagnall

, Al-Muhaizea

, Manzur

. Feeding and speech difficulties in typical congenital nemaline myopathy. Dev Med Child Neurol. 2006; 48: 29.

17.

van Gelder

, van Capelle

, Ebbink

, Moor-van Nugteren

, van den Hout

JMP

, Hakkesteegt

, et al. Facial-muscle weakness, speech disorders and dysphagia are common in patients with classic infantile Pompe disease treated with enzyme therapy. J Inherit Metab Dis. 2012; 35(3): 505-11. doi: 10.1007/s10545-011-9404-7.

18.

Wang

, Bonnemann

, Rutkowski

, Sejersen

, Bellini

, Battista

, et al. Consensus statement on standard of care for congenital muscular dystrophies. J Child Neurol. 2010; 25(12): 1559-81. doi: 10.1177/0883073810381924.

19.

Sjogreen

, Martensson

, Ekstrom

. Speech characteristics in the congenital and childhood-onset forms of myotonic dystrophy type 1. Int J Lang Commun Disord. 2018; 1-8. doi: 10.1111/1460-6984.12370.

20.

Berggren

, Hung

, Dixon

, Bounsanga

, Crockett

, Foye

, et al. Orofacial strength, dysarthria, and dysphagia in congenital myotonic dystrophy. Muscle and Nerve. 2018; 58(3): 413-417. doi: 10.1002/mus.26176.

21.

Audag

, Goubau

, Toussaint

, Reychler

. Screening and evaluation tools of dysphagia in children with neuromuscular diseases: a systematic review. Dev Med Child Neurol. 2017; 59(6): 591-596. doi: 10.1111/dmcn.13354.

22.

Dykstra

, Hakel

, Adams

. Application of the ICF in reduced speech intelligibility in dysarthria. Semin Speech Lang. 2007; 28(4): 301-11.

23.

Heckathorn

, Speyer

, Taylor

, Cordier

. Systematic review: non-instrumental swallowing and feeding assessments in pediatrics. Dysphagia. 2015; 31: 1-23. doi: 10.1007/s00455-015-9667-5.

24.

Franssen

, van den Engel-Hoek

. Logopedisch Instrument Neuromusculaire Aandoeningen voor kinderen. Logopedie. 2015; 89(7-8), 24-28.

25.

Nederlandse Vereniging voor Keel – Neus – Oorheelkunde en Heelkunde van het Hoofd – Halsgebied. Richtlijn Orofaryngeale dysfagie; 2016 [update 2017 february 1; cited 2018 july 10]. Available from: https://richtlijnendatabase.nl/richtlijn/orofaryngeale_dysfagie/startpagina_orofaryngeale_dysfagie.html.

26.

American Speech-Language-Hearing Association. Pediatric Dysphagia. [cited 2019; march 16 Available from: https//www.asha.org/PRPSpecificTopic.aspx?folderid=8589934965&section=Overview.

27.

Knuijt

, Kalf