The University of Florida (UF) Institutional Review Board (IRB) reviewed and approved
the study. Patient records were de-identified prior to analysis.
Study design, settings, and population
A dataset of 339 hospital encounters from 212 patients followed at the UF Center for
Movement Disorders and Neurorestoration (CMDNR) was used. The dataset included the
period from January 2011 to March 2013 and the study was conducted as a retrospective
cross-sectional data/chart review. A diagnosis code (International Classification of Diseases, Ninth Revision, Clinical Modification) was used to identify subjects with PD (ICD-9-CM code 332.0) who had a hospital encounter and this was used as criteria for inclusion
into the study. A hospital encounter was defined as an emergency room visit with subsequent
hospitalization, or any direct admission to the hospital due to any cause with the
exception of admission for elective procedures.
Data collection
An electronic health record system (EPIC) was used to obtain the basic demographics
and clinical data elements from each subject’s chart at the time of the hospital encounter,
including results from swallow evaluations and if aspiration precautions were administered.
Additionally, the reason for hospital encounter, length of hospital stay in days,
and whether a neurologist was consulted during these encounters were also documented.
LED was calculated using the following formula 13]: regular LD dose?+?LD CR dose × 0.75?+?LD × 0.33 if entacapone?+?pramipexole dose
× 100?+?ropinirole dose × 20?+?rotigotine dose × 30?+?pergolide dose × 1?+?bromocriptine
dose × 10?+?selegiline dose × 10?+?rasagiline dose × 100?+?amantadine dose × 1. The
Charlson index score, which has previously been used in studies with PD patients 14] was chosen to evaluate the magnitude of impact of each of the comorbidities on the
mortality risk 15]. The Charlson index predicts the ten-year mortality for each patient based on comorbidities.
Each condition is assigned a score of 1, 2, 3, or 6, depending on the risk of dying
associated with each one. Scores are summed to provide a total score to predict mortality.
Conditions screened by the Charlson index are cardiovascular disorders, pulmonary
diseases, cerebrovascular diseases, connective tissue diseases, liver and kidney diseases,
diabetes, dementia, cancer, and/or AIDS. We further divided patients into subgroups
including those with or without alteration in mental status according to the physicians
clinical examination note at the time of evaluation. Reasons for hospital encounters
were categorized based on the Emergency Severity Index (ESI) triage algorithm 16] which classifieds the documented reasons for hospitalization. According to this algorithm,
an ESI 1 and 2 were classified as a high risk for a prolonged length of stay, including
falls/fractures, pulmonary causes (such as chronic obstructive pulmonary disease,
cancer, pulmonary arterial hypertension, fibrosis, vascular disease, pneumonia), cardiac
issues/syncope, genitourinary infections, encephalopathy/drug-induced psychosis, cancer,
stroke, and dementia, and an ESI 3–5 was classified as low risk, and was inclusive
of general medical conditions (including weakness, pain, bleeding, or anorexia), other
medical issues (including seizures, vascular problems, or trauma), and gastrointestinal
issues.
Outcome measurements
The primary outcome was the frequency of aspiration pneumonia in a cohort of PD patients
during their hospital encounter. Also collected were the circumstances for the aspiration
event, including clinical features of the individual and environmental features such
as the occurrence of swallow evaluations and aspiration precautions.
To identify subjects who developed aspiration pneumonia during the encounter (in-hospital
aspiration) we used a diagnosis code (ICD-9-CM code 507.0, 507.8, and 997.32) and performed a manual chart review of each subject
and each hospital encounter in order to identify aspiration pneumonia events that
occurred at 48Â h or greater after the hospital encounter. The frequency of aspiration
pneumonia was calculated by dividing the number of hospital-acquired aspiration pneumonias
by the total number of hospital encounters, and this was then multiplied by 100. A
brief description of the swallowing evaluations and aspiration precautions is provided
in following paragraphs.
Clinical or bedside swallow evaluation (CSE)
The CSE was performed for all patients referred for a swallow evaluation. The assessment
included: 1. Patient awake/oriented/able to follow simple commands, 2. Ability to
produce sustained phonation and comment on vocal quality, 3. Ability to produce voluntary
cough on command, 4. Ability to produce saliva, swallow on command, and observation
of hyolaryngeal elevation, 5. Trials of ice-chips: observation of oral bolus manipulation;
chew (if present); timeliness of pharyngeal swallow; hyolaryngeal elevation; cough
after swallow; voice quality after swallow, and 6. Trials of thin, nectar-thick, honey-thick
liquids as deemed appropriate by the speech-language pathologist (SLP): Anterior labial
leakage, timeliness of pharyngeal swallow, hyolaryngeal elevation, cough after swallow,
and voice quality after swallow.
Rehabilitation barium swallow evaluation (RBS)
The decision to perform a RBS was made following a CSE and if the patient exhibited
signs or symptoms of an impaired swallowing mechanism. These signs included one or
more of the following: weak or absent volitional cough, cough after swallow, voice
change after swallow, absent volitional swallow, reduced or delayed hyolaryngeal elevation.
The RBS was performed in the fluoroscopy suite of the radiology department. Patients
were seated upright in the lateral viewing plane, and multiple boluses of thin, thick,
and solid consistencies were presented for evaluation by the SLP.
Aspiration precautions
The recommendation for aspiration precautions was potentially performed before or
possibly after the swallow evaluations (whether CSE or RBS). These were typically
administered pre-emptively to patients considered by the nurses or physicians to be
at-risk for aspiration, but they could also be administered after a swallowing study.
The recommendations ranged from nothing by mouth (NPO) in extreme cases of disordered
swallowing, or included modifications such as supervision during eating/drinking,
sitting upright in a chair, thickened liquids, crushed pills, IV administered pills/medications,
or various “dysphagia diets†(i.e., mechanical soft, puree, etc.).
Statistical analysis
A Mann–Whitney U Test was used to compare demographics and continuous variables (age, Charlson index,
disease duration, levodopa equivalent dosage, days from admission to swallow evaluations
and aspiration precautions, and length of hospital encounter) between those who developed
aspiration pneumonia versus those who did not develop aspiration pneumonia. A Chi
square test using an Exact Test was performed to compare categorical variables (gender,
HY stage, and neurology consultation) between populations and to learn whether preventive
swallowing measures had an effect on developing aspiration pneumonia during the hospital
encounter. Statistical analyses were performed using commercially available statistical
software (SPSS, version 19.0; SPSS, Inc., Chicago, Illinois) with a priori alpha levels
of???0.05.