development and validation of a new screening questionnaire for dysphagia in early stages of...

7
Development and validation of a new screening questionnaire for dysphagia in early stages of Parkinson's disease Janine A. Simons a, b, * , Urban M. Fietzek b , Annika Waldmann a , Tobias Warnecke c , Tibor Schuster d , Andr es O. Ceballos-Baumann b a Institute for Social Medicine and Epidemiology, Universitat zu Lübeck, Lübeck, Germany b Center for Parkinson's Disease and Movement Disorders, Schon Klinik München Schwabing (MSW), Munich, Germany c Department of Neurology, University Hospital of Muenster, Muenster, Germany d Institute of Medical Statistics and Epidemiology, Technische Universitat München (TUM), Munich, Germany article info Article history: Received 2 December 2013 Received in revised form 22 May 2014 Accepted 16 June 2014 Keywords: Dysphagia Parkinson's disease Swallowing disorders Patient-reported outcome questionnaire Screening/scales Movement disorders abstract Background: Dysphagia in patients with Parkinson's disease (PD) signicantly reduces quality of life and predicted lifetime. Current screening procedures are insufciently evaluated. We aimed to develop and validate a patient-reported outcome questionnaire for early diagnosis of dysphagia in patients with PD. Methods: The two-phased project comprised the questionnaire, diagnostic scales construction (N ¼ 105), and a validation study (N ¼ 82). Data for the project were gathered from PD patients at a German Movement Disorder Center. For validation purposes, a clinical evaluation focusing on swallowing tests, tests of sensory reexes, and beroptic endoscopic evaluation of swallowing (FEES) was performed that yielded a criteria sum score against which the results of the questionnaire were compared. Specicity and sensitivity were evaluated for the detection of noticeable dysphagia and for the risk of aspiration. Results: The Munich Dysphagia Test e Parkinson's disease (MDT-PD) consists of 26 items that show high internal consistency (a ¼ 0.91). For the validation study, 82 patients, aged 70.9 ± 8.7 (mean ± SD), with a median Hoehn & Yahr stage of 3, were assessed. 73% of patients had dysphagia with noticeable oropharyngeal symptoms (44%) or with penetration/aspiration (29%). The criteria sum score correlated positively with the screening result (r ¼ 0.70, p < 0.001). The MDT-PD sum score classied not noticeable dysphagia vs. risk of aspiration (noticeable dysphagia) with a sensitivity of 90% (82%) and a specicity of 86% (71%), and yielded similar results in cross-validation, respectively. Conclusions: MDT-PD is a valid screening tool for early diagnosis of swallowing problems and aspiration risk, as well as initial graduation of dysphagia severity in PD patients. © 2014 Elsevier Ltd. All rights reserved. 1. Introduction Dysphagia is a frequent symptom in Parkinson's disease (PD), occurring in up to 100% of patients in advanced disease stages [1,2]. As a negative prognostic predictor for the remaining lifetime in Parkinsonism, dysphagia leads to aspiration, aspiration pneumonia [3], malnutrition, dehydration [4], and reduces quality of life [5,6]. Pulmonary long-term complications such as aspiration pneumonia are among the most common causes of death in patients with PD [7,8]. Dysphagia is often diagnosed too late. Especially in early stages, the causal association between disease and swallowing disabilities remains unnoticed [9], which may be accounted for by the inability of caregivers and physicians to detect subtle swallowing problems and by the low self-awareness among PD patients [10,11]. Previous studies, matching PD patients' self-reports with the results of beroptic endoscopic evaluation of swallowing (FEES) or video uoroscopic swallowing study (VFSS), showed oropharyngeal dis- orders in over 50% of subjectively non-dysphagia PD patients [12,13], and silent aspiration occurred in up to 15% [11]. However, to date, standard diagnostics have not been routinely implemented. A directed therapy of parkinsonian dysphagia frequently begins after the onset of severe deglutition problems and subsequent health impairments [9,14]. As early diagnosis and adequate treatment of dysphagia may prevent further diseases and long-term * Corresponding author. Universitat zu Lübeck, Institute for Social Medicine and Epidemiology, Ratzeburger Allee 160, 23562 Lübeck, Germany. Tel.: þ49 451 500 5870. E-mail address: [email protected] (J.A. Simons). Contents lists available at ScienceDirect Parkinsonism and Related Disorders journal homepage: www.elsevier.com/locate/parkreldis http://dx.doi.org/10.1016/j.parkreldis.2014.06.008 1353-8020/© 2014 Elsevier Ltd. All rights reserved. Parkinsonism and Related Disorders 20 (2014) 992e998

Upload: andres-o

Post on 16-Feb-2017

216 views

Category:

Documents


2 download

TRANSCRIPT

Page 1: Development and validation of a new screening questionnaire for dysphagia in early stages of Parkinson's disease

lable at ScienceDirect

Parkinsonism and Related Disorders 20 (2014) 992e998

Contents lists avai

Parkinsonism and Related Disorders

journal homepage: www.elsevier .com/locate/parkreldis

Development and validation of a new screening questionnaire fordysphagia in early stages of Parkinson's disease

Janine A. Simons a, b, *, Urban M. Fietzek b, Annika Waldmann a, Tobias Warnecke c,Tibor Schuster d, Andr�es O. Ceballos-Baumann b

a Institute for Social Medicine and Epidemiology, Universit€at zu Lübeck, Lübeck, Germanyb Center for Parkinson's Disease and Movement Disorders, Sch€on Klinik München Schwabing (MSW), Munich, Germanyc Department of Neurology, University Hospital of Muenster, Muenster, Germanyd Institute of Medical Statistics and Epidemiology, Technische Universit€at München (TUM), Munich, Germany

a r t i c l e i n f o

Article history:Received 2 December 2013Received in revised form22 May 2014Accepted 16 June 2014

Keywords:DysphagiaParkinson's diseaseSwallowing disordersPatient-reported outcome questionnaireScreening/scalesMovement disorders

* Corresponding author. Universit€at zu Lübeck, InstEpidemiology, Ratzeburger Allee 160, 23562 Lübeck,5870.

E-mail address: [email protected] (J.A. Simo

http://dx.doi.org/10.1016/j.parkreldis.2014.06.0081353-8020/© 2014 Elsevier Ltd. All rights reserved.

a b s t r a c t

Background: Dysphagia in patients with Parkinson's disease (PD) significantly reduces quality of life andpredicted lifetime. Current screening procedures are insufficiently evaluated. We aimed to develop andvalidate a patient-reported outcome questionnaire for early diagnosis of dysphagia in patients with PD.Methods: The two-phased project comprised the questionnaire, diagnostic scales construction (N ¼ 105),and a validation study (N ¼ 82). Data for the project were gathered from PD patients at a GermanMovement Disorder Center. For validation purposes, a clinical evaluation focusing on swallowing tests,tests of sensory reflexes, and fiberoptic endoscopic evaluation of swallowing (FEES) was performed thatyielded a criteria sum score against which the results of the questionnaire were compared. Specificityand sensitivity were evaluated for the detection of noticeable dysphagia and for the risk of aspiration.Results: The Munich Dysphagia Test e Parkinson's disease (MDT-PD) consists of 26 items that show highinternal consistency (a ¼ 0.91). For the validation study, 82 patients, aged 70.9 ± 8.7 (mean ± SD), with amedian Hoehn & Yahr stage of 3, were assessed. 73% of patients had dysphagia with noticeableoropharyngeal symptoms (44%) or with penetration/aspiration (29%). The criteria sum score correlatedpositively with the screening result (r ¼ 0.70, p < 0.001). The MDT-PD sum score classified not noticeabledysphagia vs. risk of aspiration (noticeable dysphagia) with a sensitivity of 90% (82%) and a specificity of86% (71%), and yielded similar results in cross-validation, respectively.Conclusions: MDT-PD is a valid screening tool for early diagnosis of swallowing problems and aspirationrisk, as well as initial graduation of dysphagia severity in PD patients.

© 2014 Elsevier Ltd. All rights reserved.

1. Introduction

Dysphagia is a frequent symptom in Parkinson's disease (PD),occurring in up to 100% of patients in advanced disease stages [1,2].As a negative prognostic predictor for the remaining lifetime inParkinsonism, dysphagia leads to aspiration, aspiration pneumonia[3], malnutrition, dehydration [4], and reduces quality of life [5,6].Pulmonary long-term complications such as aspiration pneumoniaare among the most common causes of death in patients with PD[7,8].

itute for Social Medicine andGermany. Tel.: þ49 451 500

ns).

Dysphagia is often diagnosed too late. Especially in early stages,the causal association between disease and swallowing disabilitiesremains unnoticed [9], which may be accounted for by the inabilityof caregivers and physicians to detect subtle swallowing problemsand by the low self-awareness among PD patients [10,11]. Previousstudies, matching PD patients' self-reports with the results offiberoptic endoscopic evaluation of swallowing (FEES) or videofluoroscopic swallowing study (VFSS), showed oropharyngeal dis-orders in over 50% of subjectively non-dysphagia PD patients[12,13], and silent aspiration occurred in up to 15% [11]. However, todate, standard diagnostics have not been routinely implemented. Adirected therapy of parkinsonian dysphagia frequently begins afterthe onset of severe deglutition problems and subsequent healthimpairments [9,14]. As early diagnosis and adequate treatment ofdysphagia may prevent further diseases and long-term

Page 2: Development and validation of a new screening questionnaire for dysphagia in early stages of Parkinson's disease

Fig. 1. MDT-PD project phases.

J.A. Simons et al. / Parkinsonism and Related Disorders 20 (2014) 992e998 993

complications [15], reliable and valid screening instruments todetect dysphagia are clearly required.

The diagnostic standards to evaluate dysphagia instrumentallyare FEES, most often combined with the penetrationeaspirationscale (PAS [16]), as well as VFSS [17]. However, only a limitednumber of validated PD-specific screening tools in the form ofpatient questionnaires are available. The swallowing disturbancequestionnaire (SDQ [18]) focuses on the detection of aspirationepisodes, but is not designed to detect symptoms of earlydysphagia. Moreover, its limited clinimetric properties have beencriticized [19]. MDS-UPDRS Part II [20] and NMSQuest [21] onlyinclude two items that evaluate dysphagia in the context of theoverall syndrome, and have not been validated against instru-mental diagnostics.

This study's objective was to develop a patient outcome ques-tionnaire and validate it using criteria derived from standardizedclinical symptom rating scales and FEES. This novel instrument wasdesigned to serve as a tool for dysphagia risk screening, as well asfor grading dysphagia in PD patients.

2. Methods

2.1. Ethics approval

The research project was conducted according to the Declara-tion of Helsinki and good clinical practice. The ethics committee ofthe LudwigeMaximilians-Universit€at, Munich, Germany, approvedthe study protocol. Patients and caregivers were informed on thestudy objectives and procedures, and gave prior written consent.Personal information was collected at the investigational site andprotected according to local data protection law.

2.2. Study population & design

Patients were recruited from the Center for Movement Disor-ders and PD of the Sch€on Klinik, Munich, Germany, from 1 March2009 to 31 March 2011.

Project phase I included a total of 75 patients with PD and 30relatives who served as controls. The questionnaire, MunichDysphagia Test e Parkinson's disease (MDT-PD), and diagnosticscales were developed in a three-stage process (Part 1), and sub-sequently tested in a pilot study (Part 2).

Project phase II (validation study) included 82 PD patients whowere newly enrolled consecutively after selection from theadmission list of the hospital during a period of six months.

A flow chart of patient recruitment is available as supplemen-tary electronic file (Appendix 1). Fig. 1 provides an overview of theproject phases (including a total of 187 individuals).

Patient data were gathered from the neurological history and, ifnecessary, completed by respective investigations: motor functionand disease severity were examined using the unified Parkinson'sdisease rating scale (UPDRS) motor part III and the modified Hoehn&Yahr scale (HY). Short interviews on personal, local, and chrono-logical orientation, hospital stay, and everyday activities wereconducted to ensure the patients' ability to understand and fill outthe questionnaire. Results of mini-mental state examination(MMSE), Parkinson neuropsychometric dementia assessment(PANDA), or Montreal cognitive assessment (MoCA) ensured theexclusion of severe dementia.

Inclusion criteria were diagnosed PD, according to UK BrainBank criteria, and signed written consent.

Exclusion criteria were pre-diagnosed dysphagia (except forsubgroup 1 in project phase I), severe dementia (WHO-ICD-10F02.0), moderate to severe depression (F32.2/F32.3), other severeneurological diseases, or other conditions leading to dysphagia.

2.3. Project phase I e development

2.3.1. Part 1: questionnaire and diagnostic scale construction

1) In addition to literature review and expert consultation, semi-structured interviews with 10 PD patients (male ¼ 7; meanage ¼ 70.2 ± 9.4 years; median HY ¼ 3, ran 2.5e4; meanUPDRSIII ¼ 22.7 ± 10.1; mean MMSE ¼ 28.0 ± 1.6) and 10partners/children (m ¼ 2; mean age ¼ 60.4 ± 17.5 years) wereconducted to generate items for the detection of swallowingproblems specific for PD.

2) 20 PD patients (m ¼ 7; mean age ¼ 68.7 ± 4.0; median HY ¼ 3,range 2.5e4; mean UPDRSIII ¼ 24.7 ± 4.8; meanMMSE ¼ 28.5 ± 1.2) were asked to fill out the first, 46-itemversion of a draft questionnaire. Cognitive interviewing tech-niques for psychometric tests [22,23] were used to measure thequality of structure, content, comprehensibility, and usability(used approaches: items' alignment and ratings, think-aloud,testing in field). This led to item modification and reduction toa number of 39.

3) Diagnostic scales for swallowing examination, originally devel-oped for various etiologies, were modified to fit PD patients, asto date there is no gold standard to evaluate parkinsoniandysphagia. Scale parameters were selected according to level ofevidence, national guidelines for bedside swallowing screenings(e.g., water swallow test [24]), international generic protocolsfor endoscopy (e.g., PAS combined with clearance effectiveness),known predictors for aspiration [25,26], and Parkinson-specificsymptoms found in literature [27]. Each diagnostic parameterwas assessed using an ordinal scale that describes symptomseverity.

In order to compare drinking behavior of PD patients withhealthy persons, an orientation study was conducted that asked 20

Page 3: Development and validation of a new screening questionnaire for dysphagia in early stages of Parkinson's disease

J.A. Simons et al. / Parkinsonism and Related Disorders 20 (2014) 992e998994

voluntary, healthy relatives (m¼ 10;mean age¼ 61.4 ± 6.6 years) todrink 90 ml of water relatively quick without pausing. Swallowingrate and frequencyweremeasured by palpation, counting, and timeneeded. Measured values were compared to those found in litera-ture [28], and used to develop the first two of five severity levels(normal/slightly affected) of the timedwater swallow test as part ofthe clinical swallowing protocol (levels of modified 90 ml H2Oswallow test ranging from 0 to 4/normal to severely affected, alongwith the oropharyngeal or laryngeal dysphagia symptoms; i.e.,prolonged swallows/longer time needed, need to pause/abort thetest, wet voice, coughing, or breathing problems; see Appendix 2and 3).

2.3.2. Part 2: pilot study

1) 45 PD patients (m ¼ 29; mean age ¼ 73.0 ± 5.0 years; medianHY ¼ 3, range 2e5; mean UPDRSIII ¼ 23.7 ± 7.9;MMSE ¼ 26.2 ± 2.2; PANDA ¼ 19.9 ± 4.0) filled out the 39-itemdraft MDT-PD, and were assessed using standardized clinicaland FEES protocols. In order to test the construct validity ofMDT-PD, items concerning quality of life (QoL) and diseaseburden were compared to the results of the Parkinson's diseasequestionnaire (PDQ-39; sum score and items facing swallowingproblems) [29].

2) A qualitative analysis led to a second reduction of items (finalMDT-PD version ¼ 26 items): only those QoL items that showedpositive associations to beginning symptoms or health impair-ments (e.g., nutritional status) either in our study or in previousstudies [5,6] were retained without modification.

3) During clinical swallowing evaluation and/or FEES, some pa-tients showed (test) reactions that could not be definitelyattributed to one of the predefined symptom severity ratings ofthe newly developed diagnostic scales (Phase I, Part 1.3).Therefore, descriptions of grading were extended or furtherspecified along the observed spectrum of symptoms in order toallow unambiguous allocation. This was done in a dynamicprocess during the pilot study, in which physicians/therapistswere blinded to patients' answers already given duringMDT-PD.The modifications were consented by amulti-professional team.

2.4. Questionnaire design & intercultural translation

MDT-PD was developed in German, then translated into Englishand translated back again, complying with recommendations forculture-comparing questionnaires by application of decentering[30]. The opening page introduces the survey's goals and gives in-structions. Items are presented in first person, present tense,formulated in a concise manner, and easy to read. Technical vo-cabulary, negative wordings, and double negations are avoided.Two recurring unipolar rating scales are used for “agreement” and“symptom frequency”. Four consistent answer options are pre-sented to counteract any tendency towards the middle response,and are ranked numerically (value scores 0e3), except for the lastsub-scale, which has a dichotomous response format (0 ¼ yes/3 ¼ no). Time to fill out the questionnaire is about 10 min.

The final 26-item MDT-PD consists of four sub-scales:

I. Difficulty swallowing food and liquids (10 items): items areorganized by the sequence of the oropharyngo-esophagealswallowing act and potential pathologies.

II. Difficulty swallowing independent from food intake (4items): includes sialorrhea (drooling), (drug-related) xero-stomia, saliva penetration/aspiration, and difficulties duringpill intake, e.g., aggravated oropharyngeal transport,

residues, penetration/aspiration of the pill itself, or theconsistency with which they are ingested.

III. Further swallowing-specific and accompanying burden (9items): negatively influences the swallowing-related dailyroutine, and amplifies already acquired symptoms. Focusesare set on motor fluctuations and patients compensationbehavior to swallow easier or avoid the intake of specificconsistencies.

IV. Swallowing-specific health questions (3 items): for acquiringmedically relevant information on dysphagia or health risks.

2.5. Project phase II e validation study

After assessment using the MDT-PD-questionnaire, patientswere examined (i) clinically and (ii) endoscopically during theirmedication ON state. Test instructions as well as explanations forthe diagnostic ratings were given on the standardized protocols.Physicians/therapists examining the patients were unaware of thescreening results.

i) Clinical examination of all 82 patients included inspection ofstructure, sensory-reflex analyses, functional exam, and swal-lowing tests with different consistencies (see Appendix 2).

ii ) Due to early dismissals from hospital (N ¼ 5), FEES examina-tion was conducted with 77 subjects according to the FEESprotocol, including an extended procedure in the swallowingsection: a fiberoptic endoscope (Olympus EIVF, type P4,OLYMPUS GmbH, Medical Systems, Hamburg, Germany) wasused transnasally without local anesthesia; digital records weresaved both locally using the software rpSzene® (REHDER/PARTNER GmbH, Medical Technology, Hamburg, Germany) andon DVD. In addition to laryngeal penetration/aspiration [16],subtle symptoms of dysphagia, e.g., posterior bolus leakage,pharyngeal residues, or saliva accumulation were rated on thesymptom scales. All parameters (19 clinical/10 FEES) are listed inAppendix 2.

In both exams, nutrition ingestion was assessed with drinks orfoods from everyday life, in congruent and sufficient quantity of theingestion of three different consistencies and two different medicaldrugs were tested: (i) thin liquid (90 ml of spring water, dyed blue;H2O), (ii) solid food (half slice of bread with butter/vegetable spreadand crust, approx. 8*7*1 cm; BREAD), (iii) dry food (German butterchocolate cookie, diameter 5 cm; COOKIE), (iv) divisible uncoatedProLife VitaFit tablet (approx. 19*8*7 mm; TABLET), and (v) un-coated Hepa Lichtenstein placebo pill (diameter 8 mm; PILL).

2.6. Statistical methods

IBM SPSS Statistics for Windows (version 19.0, SPSS Inc., Chi-cago, Illinois, USA) and the software package R (version 2.14.1, RDevelopment Core Team 2011, Vienna, Austria) were used for sta-tistical analysis.

Mean or median, standard deviation, range, and interquartilerange were used for the description of continuous variables; ab-solute or relative frequencies were used for discrete/categoricaldata.

2.6.1. Reliability analysisStrength of internal consistency was examined by Cronbach's

alpha. To determine the strength of questionnaire item collinearity,variance inflation factors (VIF) were calculated. Kendall's tau co-efficients (t) were calculated to assess bivariate associations ofpatients' answers to the MDT-PD items (N ¼ 82).

Page 4: Development and validation of a new screening questionnaire for dysphagia in early stages of Parkinson's disease

J.A. Simons et al. / Parkinsonism and Related Disorders 20 (2014) 992e998 995

2.6.2. Dysphagia classification procedure & criteria sum scoreAs clinical and FEES evaluation of dysphagia produces a range of

oropharyngeal findings, various stages of severity can be differen-tiated. Based on a set of diagnostic criteria, patients were assignedto either ‘not noticeable dysphagia’, ‘noticeable dysphagia’, or ‘riskof aspiration’ (Table 1 and Appendix 3).

To validate the MDT-PD questionnaire sum score, a diagnosticcriteria sum score was calculated from the clinical and FEES as-sessments (based on 77 patients). Since 11 of 29 diagnostic pa-rameters were strong surrogates of at least one other diagnosticparameter, 18 criteria were included in the final version.

2.6.3. Validation procedureItem weighting & MDT-PD sum score: A linear regression

analysis was carried out to establish a “weighted MDT-PD sumscore” for the dependent variable “unweighted criteria sum score”with MDT-PD items as independent predictors. Resulting coeffi-cient estimates of MDT-PD items served as weight factors forsummarizing MDT-PD items (Appendix 4).

2.6.4. Strength of relationship between MDT-PD and criteria sumscore was analyzed using Spearman's correlation coefficient rho (r)

Diagnostic quality analysis: Cut-off values for weighted andunweighted MDT-PD sum scores were calculated for creating threepossible dysphagia classifications as part of the receiver operatingcharacteristic (ROC) analysis (based on 77 patients): initially, globaldiscrimination quality of the weighted/unweighted MDT-PD wasevaluated to classify “not noticeable dysphagia” vs. “any kind ofdysphagia” (oropharyngeal symptoms and/or aspiration risk). We

Table 1Set of diagnostic criteria.

Diagnostic parameters clinical (CL)/FEES (E) Symptom scale range

1 CL/PILLor CL/TABLET

0e2

2 CL/H2O swallow 0e4and/orCL/pharyngeal sensibility (PS) 0e2

3 CL/BREAD swallow 0e44 CL/COOKIE swallow 0e45 E/secretion management 0e46 E/bolus leakage H2O 0e47 E/bolus leakage BREAD 0e48 E/bolus leakage COOKIE 0e49 E/residues H2O (RES) 0e3

and/orE/clearance effectiveness H2O (CE) 0e4

10 E/residues BREAD (RES) 0e3and/orE/clearance effectiveness BREAD (CE) 0e4

11 E/residues COOKIE 0e3and/orE/clearance effectiveness COOKIE 0e4

12 E/residues PILL (RES) and 0e3E/clearance effectiveness PILL (CE) or 0e4E/residues TABLET (RES) and 0e3E/clearance effectiveness TABLET (CE) 0e4

13 E/leakage afterward H2O 0e414 E/leakage afterward BREAD 0e415 E/leakage afterward COOKIE 0e416 E/penetrationeaspiration scale PAS H2O 1e817 E/PAS BREAD 1e818 E/PAS COOKIE 1e8CLASSIFICATION RULE (for patient's overall assignment to 3 groups)0: Patients are assigned to classification “no dysphagia” when they fulfill less than 3 c1: Patients are assigned to classification “oropharyngeal dysphagia” when they fulfill2: Patients are assigned to classification “dysphagia with penetration/aspiration” wh

The classification of the diagnostic symptom scales can be obtained as supplementary e

further differentiated the classification and defined “not notice-able” vs. “noticeable dysphagia” and “not noticeable dysphagia” vs.“risk of aspiration”. The resampling method of leave-one-out cross-validation was applied. We calculated sensitivity/specificity andtheir 95% confidence intervals, positive/negative predictive values,likelihood ratios, as well as Youden indices.

Area-under-the-curve (AUC)-values were used for describingthe strength of the discriminatory ability of one single question-naire item as part of the three dysphagia classifications.

3. Results

3.1. Study population & MDT-PD questionnaire

The 26-item MDT-PD was evaluated in all 82 PD patients of thevalidation study, of whom 77 received both clinical and FEESevaluations. Table 2 shows group demographics; Appendix 5 givesan item overview.

3.2. Reliability properties

MDT-PD demonstrated high internal consistency (a ¼ 0.91).For six of the 26 MDT-PD items, VIF showed a strong multi-

collinearity (VIF > 4 by conservative interpretation), e.g., food re-mains (VIF 4.03), discharge (4.19), single swallowing (4.65), coughingwhile drinking (5.27), multiple swallowing (5.40), and choking onsaliva (5.74).

Bivariate correlation calculation for the 26 MDT-PD itemsrevealed 54 moderately positive correlations (t � 0.40). The items

Criterion classification (for each parameter)

Criterion value 0 Criterion value 1 Criterion value 2

¼ 0 ¼ 1 ¼ 2

H2O � 1 and PS ¼ 0 H20 ¼ 1 and PS � 1 H20 � 3or orH2O ¼ 0 H2O ¼ 2�1 ¼2 �3�1 ¼2 �3¼0 ¼1 or 2 �3�1 ¼2 �3�1 ¼2 �3�1 ¼2 �3RES ¼ 0 RES¼ 1 and CE ¼ 2 RES ¼ 3or orRES ¼ 1 and CE � 2 RES ¼ 2RES ¼ 0 RES ¼ 1 and CE > 2 RES ¼ 3or orRES ¼ 1 and CE � 2 RES ¼ 2RES ¼ 0 RES ¼ 1 and CE > 2 RES ¼ 3or orRES ¼ 1 and CE � 2 RES ¼ 2RES ¼ 0 RES ¼ 1 and CE > 2 RES ¼ 3or orRES ¼ 1 and RES ¼ 2CE � 2�1 ¼2 �3�1 ¼2 �3�1 ¼2 �3¼1 ¼2 �3¼1 ¼2 �3¼1 ¼2 �3

riteria of value 1at least 3 criteria of value 1en they fulfill at least 2 criteria of value 2

lectronicefile (Appendix 3).

Page 5: Development and validation of a new screening questionnaire for dysphagia in early stages of Parkinson's disease

Table 2Group demographics.

Study cohort N ¼ 82Patients of validationstudy including 5 dropeouts(Complete neurological screening,MDT-PD questionnaire, clinicalswallowing examination, only77/82 endoscopic swallowingexamination FEES)

N ¼ 77Patients of validation study(Complete neurological screening,MDT-PD questionnaire, clinical swallowing examination andendoscopic swallowing examinationFEES)

Patient characteristics N (%) or mean (SD), minemax N (%) or mean (SD), minemax

Gender Female 36 (43.9%) 33 (42.9%)Male 46 (56.1%) 44 (57.1%)

School education <12 years 49 (59.8%) 45 (58.4%)�12 years 11 (13.4%) 11 (14.3%)Academic degree 22 (26.8%) 21 (27.3%)

Modified Hoehn & Yahr scale (0e5) 2 6 (7.3%) 6 (7.8%)2.5 11 (13.4%) 10 (13.0%)3 32 (39.0%) 29 (37.7%)4 30 (36.6%) 29 (37.7%)5 3 (3.7%) 3 (3.9%)

Age 70.85 (8.74), 49e90 70.47 (8.40), 49e87Disease durationa 11.01 (6.31), 0e30 11.19 (6.27), 0e30UPDRSIII 0e108 29.45 (13.26), 7e58 29.59 (13.32), 7e58Cognitive state MoCA (0e30) 23 (5.82), 2e30 22.8 (5.87), 2e30

MMSE (0e30) 27 (3.65), 19e30 27 (3.38), 19e30PANDA (0e30) 16.4 (2.3), 13e19 16.5 (2.29), 13e19

BMI 24.3 (3.77), 16.3e33.1 24.3 (3.72), 16.3e32.2

a Initial diagnosis; UPDRS pt. III, unified Parkinson's disease rating scale, Part III motor examination; MoCA, montreal cognitive assessment; MMSE, mini-mental stateexamination; PANDA, Parkinson neuropsychometric dementia assessment; BMI, body mass index.

J.A. Simons et al. / Parkinsonism and Related Disorders 20 (2014) 992e998996

coughing while drinking and coughing while eating showed thestrongest correlation (t ¼ 0.62). A correlogram of MDT-PD items isavailable (Appendix 6).

3.3. Dysphagia classification

Patients were classified as showing “no dysphagia” (N ¼ 21,27.3%), “oropharyngeal dysphagia” (N ¼ 34, 44.2%), or “dysphagiawith penetration/aspiration” (N¼ 22, 28.6%). They obtained ameancriteria sum score of 19.75 (min/max ¼ 5e51) and a median-weighted MDT-PD sum score of 4.85 (interquartilerange ¼ 2.95e6.96). For evaluation of skewness, floor and ceilingeffects see Appendix 7.

3.4. Validity properties

The higher patients scored in clinical and FEES evaluations withregard to criteria sum score, the higher they rated their swallowingsymptoms in the weighted MDT-PD (r ¼ 0.70, p < 0.001).

3.4.1. Diagnostic validity & cross-validityWeighting theMDT-PD sum score resulted in a higher predictive

ability compared to the unweighted score. The discriminatoryabilities of the weighted MDT-PD with regard to the three

Table 3MDT-PD diagnostic quality.

Dysphagia groups Not noticeable (0) vs. noticeable (1) Not noticeable

Original Cross-validation Original

Sens (CI) 82.4% (0.696e0.952) 82.4% (0.696e0.952) 90.0% (0.775eSpec (CI) 71.4% (0.521e0.907) 61.9% (0.411e0.827) 85.7% (0.707ePPV/NPV 82.4%/71.4% 77.8%/68.4% 87.0%/90.0%YI 0.54 0.44 0.76LRþ 2.9 2.2 6.3Cut off 3.65 3.63 4.79

Sensitivity (Sens), specificity (Spec), confidence interval (CI), positive/negative predictive

dysphagia classifications “not noticeable dysphagia” vs. 1)“noticeable dysphagia”, 2) “risk of aspiration”, or 3) “any kind ofaspiration”, resulted in good to very good characteristic values(Table 3).

3.4.2. Single-item discrimination validityIn both the cases “not noticeable” vs. 1) “noticeable dysphagia”

and 2) vs. “risk of aspiration”, the AUC values of 25 of 26 items had,ranged from 0.512 to 0.876 (median 0.634/0.695), thus lying abovethe threshold of chance probability (0.5). All 26 items fulfilled thiscriterion when the goal was to distinguish “not noticeabledysphagia” vs. 3) “any kind of dysphagia” (0.644). In addition, thelower borders of the 95% confidence interval of 10, 15, and 16 items,for the classification groups 1), 2), and 3) respectively, exceeded thevalue 0.5 by themselves (range 0.507e0.765, median 0.528/0.56/0.566; upper border: range 0.762e0.986, median 0.818/0.873/0.811).

3.5. Web-based evaluation of results

For facilitating routine application, a cost- and license-free webapplication (for non-commercial clinical use) was developed thatworks on all common operating systems and devices. The programprovides a clinical interpretation of the resulting MDT-PD sum

(0) vs. risk of aspiration (2) Not noticeable (0) vs. any dysphagia (3)

Cross-validation Original Cross-validation

1.025) 90.0% (0.775e1.025) 85.2% (0.760e0.944) 85.2% (0.760e0.944)1.007) 81.0% (0.642e0.978) 71.4% (0.521e0.907) 66.7% (0.465e0.869)

81.8%/89.5% 88.5%/65.2% 86.8%/63.6%0.71 0.57 0.524.7 3.0 2.64.75 3.65 4.75

value (PPV/NPV),Youden index (YI), likelihood ratioþ (LRþ); cut-off point (Cut off).

Page 6: Development and validation of a new screening questionnaire for dysphagia in early stages of Parkinson's disease

J.A. Simons et al. / Parkinsonism and Related Disorders 20 (2014) 992e998 997

score (allocated to the classifications “no dysphagia”, “oropharyn-geal dysphagia”, and “dysphagia with risk of aspiration” e listed inAppendix 5). Both MDT-PD questionnaire and web application canbe accessed via www.mdt-parkinson.de.

4. Discussion

In order to detect dysphagia PD patients at the onset of theirsymptoms, we developed a novel patient-reported outcome ques-tionnaire with the goal to achieve high sensitivity while guaran-teeing appropriate clinical specificity. Thus, MDT-PD enables toscreen for dysphagia symptoms that might otherwise go unnoticed.As we aimed to detect beginning symptoms, comparatively moreoropharyngeal disabilities were observed in one or more nutri-tional consistencies than episodes of penetration or aspiration, andthese subtle symptoms later leading to penetration/aspirationwereeven the most frequent findings. PD staging and PD duration ofinvestigated patients were similar to those of comparable cohorts[18,27].

In patients who are at risk of laryngeal aspiration, MDT-PD de-tects dysphagia with high sensitivity and specificity. As aspirationpneumonia is a severe complication of dysphagia and a significanthealth threat for patients with PD, this instrument could assistclinical decision making, e.g., on whether a patient should be keptnil by mouth until the clinical picture and severity of dysphagia arecomprehensively assessed.

One of the major strengths of our study is that the overallsample of 187 individuals tested to develop and validate the MDT-PD questionnaire is considerably larger than in comparable vali-dation studies. Compared with the existing SDQ questionnaire [18],MDT-PD includes typical aspects of swallowing-specific burden andParkinsonism, such as motor fluctuations, and has the advantage toidentify even early oropharyngeal symptoms of impaired swal-lowing that could be further evaluated by clinical and endoscopicdiagnostics: for MDT-PD validation 94% of participants underwentFEES additionally to clinical tests; in contrast, FEES examinations ofonly those patients who reported swallowing disturbances mightcause a selection bias. The selection of only one single diagnosticparameter (penetration/aspiration) to validate the SDQ question-naire makes it seem inappropriate for revealing early stages ofdysphagia. To assess the full range of dysphagia from beginningsymptoms to severe risk of aspiration, 18 gradually rated parame-ters (set of criteria) were used to validateMDT-PD. As research trialshad shown that, in contrast to earlier studies, it was absolutelynecessary to test patients with several pieces or swallows of solidfood and a minimum of 90 ml of water to be able to adequatelyidentify pharyngeal residues, penetration or aspiration, (especiallysilent aspiration) this finding [26,31] was applied for MDT-PD[26,31], to avoid that the risk of penetration/aspiration is under-estimated. To counteract any overoptimistic evaluations based onthis study sample, we additionally verified the resulting classifi-cation values by cross-validation. Moreover, the usage of positive/negative predictive values and likelihood ratios, respectively, tointerpret the dysphagia classifications enables the clinician tostratify patients according to dysphagia severity in medical prac-tice. The user-friendly web application offers quick computation ofthe weighted MDT-PD sum score to avoid time-consuming manualcalculation.

FEES were carried out within five days after clinical examina-tion. On one hand, this delay might have produced inconsistenciesof data measurement. On the other hand, for PD patients, a delay offive days can make the same clinical difference as a significantlyshorter delay of 3 h, as long as both swallowing evaluations aredone in the same ON/OFF state of their medication cycle. Due tomotor fluctuations [32] as well as day-to-day variation, PD patients

are best evaluated at several time points in order to better detectswallowing problems. All study patients were examined during theON phase of their medication cycle. Median time after medicationintake for clinical/endoscopic examinations was 120/90 min. OFFphase evaluations would probably yield differing results, and futureevaluations of dysphagia should account for varying symptomsaccording to the levodopa cycle.

We have not yet determined the test-retest reliability of MDT-PD: we expected biased results during a second survey, becausepatients were informed on how they passed the clinical andendoscopic swallowing examinations. Inter-rater reliability of thestandardized swallowing diagnosis procedure has been deter-mined, though, and will be the subject of another publication.

Whether MDT-PD can produce reliable results for patients withsecondary or atypical parkinsonian syndromes has to be investi-gated in future studies. Similarly, as patients with severe dementiaand/or chronic depression were excluded from the study, usersshould be aware of the fact that reduced cognitive abilities anddepressed mood might alter answer behavior and hinder a validinterpretation of results.

In conclusion, we present a valid screening tool for evaluation ofearly disease stages, for detection of dysphagia with beginningoropharyngeal symptoms, and for assessing dysphagia severity inadvanced patients at risk of chronic aspiration. Depending on thetest results, continuous swallowing assessments as well as appro-priate therapy, dietary adaptation, or compensatory strategiesshould be employed for the conservation of QoL and prevention ofhealth risks.

Financial disclosure/conflict of interest/funding sources

There was no financial support for the study or other fundingsources involved. The following competing interests are to reportconcerning the research related to this manuscript: Dr. Fietzek wasin part supported by an unrestricted educational grant of theGerman Parkinson Association (Deutsche Parkinson Vereinigunge.V.). Dr. Simons, Dr. Waldmann, PD Dr. Warnecke, Dr. Schuster andProf. Dr. Ceballos-Baumann have no financial disclosures with re-gard to this work.

Full financial disclosures of all authors for the past 3 years

Dr. Simons: nothing to report.Dr. Fietzek has received honoraria for speeches from Desitin,

MEDA Pharma, Ipsen, Merz Pharmaceuticals, and Pharmallergan.He is member of an advisory board on botulinum toxin treatmentfor Ipsen Pharma. His employment was supported with a grantfrom the German Parkinson Association (Deutsche Parkinson Ver-einigung e.V.) and the German Foundation for Neurology (DeutscheStiftung Neurologie).

PD Dr. Waldmann: nothing to report.PD Dr. Warnecke has received honoraria for advisory work or

speeches from Archimedes, Abbvie, Teva, UCB.Dr. Schuster's current affiliation is supported by the Canadian

Network for Observational Drug Effect Studies (CNODES). CNODESis funded by a grant from Health Canada, the Drug Safety andEffectiveness Network (DSEN), and the Canadian Institutes forHealth Research (CIHR).

Prof. Dr. Ceballos-Baumann has received an unrestricted grantfrom the German Parkinson Association (Deutsche Parkinson Ver-einigung e.V.) and the German Foundation for Neurology (DeutscheStiftung Neurologie). He has received honoraria for advisory workor speeches from Archimedes, Allergan, Desitin, GlaxoSmithKline,Ipsen, Licher-MT, Meda, Merz, Medtronic, Novartis, Orion, Teva,UCB.

Page 7: Development and validation of a new screening questionnaire for dysphagia in early stages of Parkinson's disease

J.A. Simons et al. / Parkinsonism and Related Disorders 20 (2014) 992e998998

Author roles

Dr. Simons: research project (conception, organization, execu-tion), statistical analyses (design, execution), manuscript (writingof 1st draft).

Dr. Fietzek: research project (conception), statistical analyses(review and critique), manuscript (review and critique).

PD Dr. Waldmann: statistical analyses (review and critique),manuscript (review and critique).

PD Dr. Warnecke: manuscript (review and critique).Dr. Schuster: statistical analyses (design, execution), manuscript

(review and critique).Prof. Dr. Ceballos-Baumann: manuscript (review and critique).

Acknowledgments

We are grateful to S. Eder, M.Sc. Inf., for his close collaborationwith regard to web application programming, to E. Wagner-Sonntag, SLP Ma., for consultations in stages of development andrealization, to the otolaryngologists of the MSW clinic for endo-scopic consultations, as well as for the support of speech-language-swallowing pathologists, neurologists, neuropsychotherapists, andphysiotherapists. Special thanks belong to all participating MSWpatients and their relatives as well as to the management of theMSW clinic for giving their permission to use local facilities andutilities.

Appendix A. Supplementary data

Supplementary data related to this article can be found at http://dx.doi.org/10.1016/j.parkreldis.2014.06.008.

References

[1] Kalf J, de Swart B, Bloem B, Munneke M. Prevalence of oropharyngealdysphagia in Parkinson's disease: a meta-analysis. Parkinsonism Relat Disord2012;18:311e5.

[2] Miller N, Noble E, Jones D, Burn D. Hard to swallow: dysphagia in Parkinson'sdisease. Age Aging 2006:614e8.

[3] Pfeiffer RF. Gastrointestinal dysfunction in Parkinson's disease. Lancet Neurol2003:107e16.

[4] Barichella M, Cereda E, Pezzoli G. Major nutritinal issues in the managementof Parkinson's disease. Mov Disord 2009;24:1881e92.

[5] Plowman-Prine E, Sapienza C, Okun M, Pollock S, Jacobson C, Wu S, et al. Therelationship between quality of life and swallowing in Parkinson's disease.Mov Disord 2009;24:1352e8.

[6] Leow L, Huckabee ML, Anderson T, Beckert L. The impact of dysphagia andquality of life and Parkinson's disease as measured by quality of life (SWAL-QOL) questionnaire. Dysphagia 2010;25:216e20.

[7] Müller J, Wenning G, Verny M, McKee A, Chaudhuri KR, Jellinger K, et al.Progression of dysarthria and dysphagia in postmortem-confirmed parkin-sonian disorders. Arch Neurol 2001;58:259e64.

[8] Hely M, Morris J, Traficante R, Reid W, O'Sullivan D, Williamson P. The sydneymulticentre study of Parkinson's disease: progression and mortality at 10years. J Neurol Neurosurg Psychiatry 1999;67:300e7.

[9] Volonte MA, Porta M, Comi G. Clinical assessment of dysphagia in early phasesof Parkinson's disease. Neurol Sci 2002;23:121e2.

[10] Miller N, Allcock L, Hildreth AJ, Jones D, Noble E, Burn DJ. Swallowing prob-lems in Parkinson disease: frequency and clinical correlates. J Neurology,Neurosurg Psychiatry 2009:1047e9.

[11] Ali GN, Wallace KL, Schwarz R, DeCarle DJ, Zagami AS, Cook IJ. Mechanism oforal-pharyngeal dysphagia in patients with Parkinson's disease. Gastroen-terology 1996:383e92.

[12] Bird MR, Woodward MC, Gibson EM, Phyland DJ, Fonda D. Asymptomaticswallowing disorders in eldery patients with Parkinson's disease: a descrip-tion of findings on clinical examination and videoflouroscopy in sixteen pa-tients. Age Aging 1994:251e4.

[13] Fuh J, Lee R, Wang S, Lin C, Wang P, Chiang J, et al. Swallowing difficulty inParkinson's disease. Clin Neurol Neurosurg 1997;99:106e12.

[14] Monte FS, Da Silva-Junior FP, Braga-Neto P, Nobre e Souza MA, de Bruin VM.Swallowing abnormalities and dyskinesia in Parkinson's disease. Mov Disord2005;20:457e62.

[15] Troche MS, Okun MS, Rosenbek JC, Musson N, Fernandez HH, Rodriguez R,et al. Aspiration and swallowing in Parkinson disease and rehabilitation withEMST: a randomized trial. Neurology 2010:1912e9.

[16] Rosenbek JC, Robbins JA, Roecker EB, Coyle JL, Wood JL. A penetration-aspi-ration scale. Dysphagia 1996:93e8.

[17] Langmore S, Schatz K, Olson N. Endoscopic and videofluoroscopic evaluationsof swallowing and aspiration. Ann Otol Rhinol Laryngol 1991;100:678e81.

[18] Manor Y, Giladi N, Cohen A, Fliss D, Cohen J. Validation of a swallowingdisturbance questionnaire for detecting dysphagia in patients with Parkin-son's disease. Mov Disord 2007;22:1917e21.

[19] Evatt ML, Chaudhuri KR, Chou KL, Cubo E, Hinson V, Kompoliti K, et al. Dys-autonomia rating scales in Parkinson's disease: sialorrhea, dysphagia, andconstipation - critique and recommendations by movement disorders taskforce on rating scales for Parkinson's disease. Mov Disord 2009;24:635e46.

[20] Goetz CG, Fahn S, Martinez-Martin P, Poewe W, Sampaio C, Stebbins GT, et al.Movement disorder society-sponsored revision of the unified Parkinson'sdisease rating scale (MDS-UPDRS): process, format, and clinimetric testingPlan. Mov Disord 2006;22:41e7.

[21] Chaudhuri KR, Martinez-Martin P, Brown RG, Sethi K, Stocchi F, Odin P, et al.The Metric properties of a novel non-motor symptoms scale for Parkinson'sdisease: results from an international pilot study. Mov Disord 2007;22:1901e11.

[22] Schwarz N. Cognitive aspects of survey Methodology. Appl Cogn Psychol2007;21:227e87.

[23] Schwarz N, Sudmann S. Answering questions: methodology for determiningcognitive and communicative process in survey research. San Francisco: Jos-sey-Bass; 1996.

[24] Suiter DM, Leder SB. Clinical utility of the 3-ounce water swallow test.Dysphagia 2008:244e50.

[25] Murray J, Langmore SE, Ginsberg S, Dostie A. The significance of accumulatedoropharyngeal secretions and swallowing frequency in predicting aspiration.Dysphagia 1996:99e103.

[26] Eisenhuber E, Schima W, Schober E, Pokieser P, Stadler A, Scharitzer M, et al.Videoflouroscopic assessment of patients with Dysphagia. Pharyngeal reten-tion is a predictive factor for aspiration. Am J Roentgenol 2002;178:393e8.

[27] Lam K, Kwai Y, Lam F, Kwong Lau K, Kay Chan Y, Yee Ling KE, et al. Simpleclinical tests may predict severe oropharyngeal dysphagia in Parkinson'sdisease. Mov Disord 2007;22:640e4.

[28] Nathadwarawala K, Nicklin J, Wiles C. A timed test of swallowing capacity forneurological patients. J Neurol Neurosurg Psychiatry 1992;55:822e5.

[29] Jenkinson C, Fitzpatrick R, Peto V, Greenhall R, Hyman N. The Parkinson'sdisease questionnaire(PDQ-39): development and validation of a Parkinson'sdisease summary index score. Age Ageing 1997:353e7.

[30] Smith T. Developing and evaluating cross-national survey instruments. In:Sea Presser, editor. Methods for testing and evaluating survey questionnaires.Hoboken, NJ: Wiley; 2004. pp. 431e52.

[31] Leder SB, Suiter DM, Green BG. Silent aspiration risk is volume-dependent.Dysphagia 2011;26:304e9.

[32] Bateson M, Gibbert F, Wilson R. Salivary symptoms in Parkinson's disease.Arch Neurol 1973;29:274e5.