Introduction

Dysphagia is a serious yet common problem in older adults, especially for those in hospital settings and in subacute care facilities (i.e., skilled nursing, rehabilitation, and long-term care facilities) [1–3]. Patients with stroke or hip fracture, the most common reasons for subacute care admission [4, 5], are at especially high dysphagia risk. From 40 to 70% of older adults with acute stroke experience dysphagia [6–9]. Moreover, hip fracture patients discharged to subacute care have high rates of coexisting dementia [10–12], which places them at significantly increased dysphagia risk [13–16]. Dysphagia leads to a myriad of complications, including malnutrition, dehydration, and pneumonia, costing more than $4.4 billion annually [17, 18]. It is often diagnosed within the hospital setting by speech-language pathologists (SLP), who assess swallowing ability and make effective dietary, behavioral, and provider recommendations to decrease the risk of dysphagia-related complications [19–25]. However, hospital-based physicians and SLPs rarely accompany patients to the post-hospital care setting [26], and post-hospital communication of patient care plans is often problematic [26–30]. Poor discharge communication could lead to inappropriate post-hospital dysphagia care, with resultant aspiration pneumonia and need for costly rehospitalization.

The hospital discharge summary is the only document mandated by The Joint Commission to convey the patient’s care plan to the post-hospital setting [31]. Although hospitals often utilize additional discharge paperwork, these other documents are institution-specific, not required, and not always present [27, 32–35]. Direct verbal communication between care settings is rare [27]. Despite the critical communication role discharge summaries play, they are not standardized and often lack important components that experts recognize as crucial to ensuring patient safety [27, 28, 30, 36]. It remains unknown how well discharge summaries communicate SLP dysphagia recommendations to post-hospital settings.

To enhance the design of transitional care programs that improve between-facility communication, we examined the rate of SLP dysphagia recommendation omissions in hospital discharge summaries for stroke and hip fracture patients transitioning from hospital to subacute care facilities.

Methods

Study Sample

We identified all hospitalized patients 18 years and older with primary diagnoses of stroke or pelvis/hip/femur fracture who received a billed inpatient SLP dysphagia evaluation and who were discharged to subacute care facilities during the period 2003–2005 from a single large academic medical center. We established primary diagnoses using the International Classification of Diseases, 9th edition (ICD-9) diagnosis code in the first position on the acute hospitalization discharge diagnosis list. ICD-9 codes of 431, 432, 434, and 436 were used to identify stroke [37–39], and 805.6, 805.7, 806.6, 806.7, 808, and 820 were used to identify pelvis/hip/femur fracture (hereafter simply “hip fracture”) [40–42]. We identified discharges to subacute care facilities through the use of administrative data compiled on a mandatory basis for all study hospital patients by hospital case managers. Internal testing of these data by the study hospital noted greater than 95% reliability of this discharge field. We identified patients with inpatient SLP evaluations (either bedside or instrumental) by examining hospital billing records for Current Procedural Terminology (CPT) codes of 92610 (“evaluation of oral and pharyngeal swallowing function”), 92611 (“motion fluoroscopic evaluation of swallowing”), and 92612 (“flexible fiberoptic endoscopic evaluation of swallowing”) billed out of the study hospital’s swallowing service. The initial sample size was 218 prior to exclusions.

SLP hospital chart notes for each patient were located within the combination paper/electronic patient hospital chart for the eligible hospitalization. Discharge summaries for all eligible patients were obtained electronically from the study hospital. Patients were excluded if they did not have a discharge summary (N = 2), did not have dysphagia recommendations listed in their SLP hospital chart notes (N = 10), were discharged to hospice or comfort care (N = 1), if it was clear from their discharge summary that they did not have a diagnosis of stroke or hip fracture (N = 12), or were not discharged to a subacute care facility (N = 6), for a final sample size of 187. No patient was included more than once in the sample. The Institutional Review Board (IRB) at the participating university approved this study with a waiver of consent.

Dysphagia Recommendation Categorization

We developed a coding scheme for all recommendations typically made by a SLP during the routine course of dysphagia evaluation and treatment. To accomplish this, we convened a consensus team of two SLPs, two physicians, and one medical student (the authors) to locate typical SLP recommendations via a review of the dysphagia literature [43–52] and to create a logical categorization of all recommendations found (N = 165). The team created seven major categories of dysphagia recommendations, including (1) Dietary Recommendations and Restrictions, (2) Postural and Compensatory Techniques, (3) Rehabilitative Techniques, (4) Pacing, Sizing, and Procedural Techniques, (5) Medications—Pill Recommendations, (6) Care Provider and Communication Recommendations, and (7) Environment/Other (Table 1). Large categories were divided into subcategories. For each specific recommendation within each category/subcategory, we applied a distinct 4-digit code that was used in the coding and analysis processes.

Table 1 Categorizations and subcategorizations of common recommendations made by speech-language pathologists (SLP) during inpatient dysphagia evaluations
Full size table

Abstraction and Coding Process

Final SLP Hospital Chart Note

Through a manual review of all documentation from each patient’s eligible hospitalization, the last SLP note containing recommendations prior to discharge (i.e., the “final SLP note”) was identified. Recommendations within this note were abstracted verbatim into electronic forms by a single medical abstractor (medical student) using a standardized abstraction protocol and manual. (Prior to formal chart abstraction activities, this abstractor underwent half a day of training on study protocol and abstraction approaches, including test abstractions and parallel abstractions with immediate feedback.) Each abstracted recommendation was then coded using the 4-digit codes developed above. To assess the reliability and validity of this process, a SLP who was originally involved in 5% of our sample’s care and who had performed and written the final SLP notes on these patients herself, performed retrospective reabstractions of her dysphagia recommendations within all of her own final notes. She was blinded to the original abstraction results. These reabstractions were coded and compared with the original abstractions. A total of 66 SLP dysphagia recommendations were compared, with a total agreement of 99% between the two abstractors (Cohen’s κ = 0.9).

Hospital Discharge Summary

Two trained medical abstractors (one nurse practitioner and one physician), using standardized abstraction protocols, forms, and manuals, reviewed all sample discharge summaries for the presence or absence of dysphagia recommendations/orders. All dysphagia recommendations within the discharge summaries were abstracted verbatim onto paper abstraction forms, entered into an electronic database, and manually coded. Ten percent of discharge summaries were reabstracted with a 92% interabstractor agreement noted for the presence/absence of dysphagia recommendations (Cohen’s κ = 0.7). The discharge summary abstraction team was fully blinded to all contents of the SLP hospital chart notes.

Analysis

We calculated the prevalence of dysphagia recommendations within final SLP hospital chart notes and discharge summaries. Next, for each patient, we compared the coded dysphagia recommendations obtained from the patient’s final SLP hospital chart note with those obtained from the patient’s discharge summary. Discharge summary omissions of specific SLP dysphagia recommendations were noted for each patient. Omission frequencies were calculated for each dysphagia recommendation category and subcategory. Analyses were performed using SAS version 9.1 and STATA version 10.1 [53, 54].

Results

Patient and Discharge Summary Characteristics

Of the 187 eligible patients within this study, 159 (85%) had a primary diagnosis of stroke while 28 (15%) had a primary diagnosis of hip fracture. Discharge summaries averaged 3.6 pages (range = 2–9) and originated from a variety of hospital services, including neurosurgery, neurology, orthopedic surgery, and general internal medicine. Nearly all of the discharge summaries were dictated by a physician resident (e.g., medical resident, surgical resident, neurology resident), although 96% were ultimately reviewed, edited, and signed by the attending physician.

Prevalence of Dysphagia Recommendations

Final SLP hospital chart notes contained an average of 5.6 recommendations per note (range = 1–15), while patient discharge summaries contained an average of 1.4 recommendations per discharge summary (range = 0–9). Both SLP notes and discharge summaries included “dietary recommendations and restrictions” most often, with 99% of final SLP notes and 52% of discharge summaries including at least one recommendation within this category (Fig. 1). “Care provider and communication recommendations” were the next most often included, with 65% of SLP notes and 18% of discharge summaries including at least one recommendation within this category. “Postural and compensatory techniques” were the third most often included in both note types, followed by the categories of “pacing, sizing, and procedural techniques,” “medications—pill recommendations,” “rehabilitative techniques,” and “environment.” Prevalence of the most common specific recommendations within each category is demonstrated in Appendix Table 3. Overall, dysphagia recommendations were less often included within discharge summaries than within SLP notes, regardless of the category.

Fig. 1
figure 1

Prevalence of dysphagia recommendations, by category, within final speech-language pathologist (SLP) hospital chart notes and within discharge summaries for stroke and hip fracture patients discharged to subacute care facilities (N = 187). Light gray SLP note, dark gray discharge summary

Full size image

Omission of Dysphagia Recommendations Within Discharge Summaries

Table 2 demonstrates the frequencies at which patient discharge summaries omitted specific SLP dysphagia recommendations. Overall, 45% of patient discharge summaries omitted all of the dysphagia recommendations made within the final SLP hospital note, while 42% of discharge summaries omitted at least one (but not all) of the SLP recommendations (i.e., omitted some recommendations). Thirteen percent of patient discharge summaries included all of the SLP recommendations made (i.e., omitted no recommendations).

Table 2 Discharge summary omissions of dysphagia recommendations made by speech-language pathologists (SLP) for stroke and hip fracture patients discharged to subacute care facilities (N = 187)
Full size table

Forty-seven percent of patients with dietary recommendations and restrictions made by their SLP had these recommendations completely omitted from their discharge summary (Table 2). This category had the lowest omission rate of all categories studied. In this category, SLP tube-feeding recommendations were the least commonly omitted. All other types of SLP dietary recommendations were omitted at rates of 54% or greater. Recommendations for diets other than “general” accounted for approximately 60% of all omissions within the food recommendation category, while recommendations for liquid consistencies other than “thin” accounted for approximately 22% of all omissions within the liquid recommendation category (Appendix Table 3).

From 79 to 100% of patients with nondietary SLP recommendations had these other recommendations fully omitted from their discharge summaries (Table 2). The most numerous specific omissions in these nondietary categories included recommendations for elevating the head of the patient’s bed during or after meals, one-to-one supervision or feeding assistance during meals, eating slowly or with only small bites, performing a chin tuck during swallowing, crushing tablet medications, specific rehabilitative tongue/mouth exercises, and instructions for following up with the SLP for further evaluation (Appendix Table 3).

Rarely, discharge summaries included specific dysphagia recommendations not made within the SLP hospital chart note (Appendix Table 3). The most common of these recommendations were instructions for pureed or mechanical soft diets (20 discharge summaries) and for one-to-one feeding assistance (7 discharge summaries).

Discussion

In this study we found that inpatient SLP dysphagia therapy recommendations were frequently omitted from the discharge summaries of subacute care patients at high risk for aspiration pneumonia. Nondietary recommendations were omitted at the highest rates, in some categories nearing 80-100%, while dietary recommendations were omitted for nearly half of all patients. To our knowledge, this is the first study to examine and report on deficiencies in dysphagia therapy communication at the time of hospital discharge.

The frequent omission of dysphagia recommendations in hospital discharge summaries may be attributable to a number of dictating provider (i.e., physician) and system factors. Despite the demonstrated effectiveness of dysphagia therapies in preventing aspiration and subsequent pneumonias [19, 20, 23–25], physicians may undervalue the importance of these therapies, preferentially focusing on physician-prescribed therapies (i.e., medications) for transmission within the discharge summary plan. The communication of dysphagia therapies may also be perceived by the dictating physician as a “nursing role,” one that is dealt with during the nursing hand-off (i.e., the telephone communication which typically occurs between the discharging hospital nurse and the receiving subacute care nurse at the time of the patient’s discharge from the hospital). However, nursing hand-offs do not result in written documentation as universally present or as widely disseminated as the discharge summary [27]. In addition, system factors, including poor in-hospital communication [55–57], large workloads [58, 59], and cumbersome discharge summary and medical record systems [60–63], likely contribute to dysphagia omissions.

This is not the first study to demonstrate omission of critical patient care plan components in hospital discharge summaries. Studies of discharge summaries in Britain and Canada have demonstrated frequent omissions of important details [27, 35, 36, 64–67]. A systematic review by Kripalani et al. [27] noted that discharge summaries frequently omit diagnostic test results, treatment courses, discharge medications, pending test results, and follow-up plans. However, there is a notable lack of attention to treatment plan components made by allied health providers in these studies, including dysphagia treatment recommendations made by hospital-based SLPs. As hospitalized older adults increasingly rely on multidisciplinary care teams and as research continues to highlight the critical impact transitional care quality has on patient safety in the early post-discharge period [68–71], it becomes clear that a shift in the physician-centered approach to discharge summary documentation may be needed.

Although patient outcomes were not studied within this particular analysis, the potential impact that dysphagia omissions may have on patient health is concerning. The evidence-based dysphagia therapy recommendations made by inpatient SLPs have been shown to decrease adverse patient events [19, 20, 23–25]. However, if these therapies are not communicated to or continued within the post-hospital care setting, any benefits they may have conveyed could be lost. As such, omission of food and body-positioning recommendations within discharge summaries may lead to inappropriate or unsafe patient care, thus increasing the risk of aspiration and subsequent pneumonia within the subacute care facility. This is an important linkage because accreditation and quality agencies rarely focus on the specific content of discharge summaries, concentrating instead upon the mere presence or absence of the signed document [31]. Future studies that strengthen the connection between the quality of dysphagia therapy communication at the time of hospital discharge and dysphagia-specific patient outcomes are needed.

This study has some limitations that should be considered. The retrospective nature of this analysis makes it impossible for us to determine if some dysphagia therapies recommended by SLPs were purposefully omitted from discharge summaries by physicians who felt these recommendations were not appropriate for the patient at the time of hospital discharge. However, the remarkably high omission rate of dysphagia recommendations within discharge summaries and the high incidence of long-term dysphagia in subacute care populations [1, 13–15] make it unlikely that purposeful omissions explain the bulk of these findings. Second, this study was conducted in a single, large academic medical center in which most discharge summaries are authored by physician residents and in which most stroke patients are cared for within a dedicated stroke unit. This may limit the generalizability of our findings, especially considering that academic discharge summaries may differ from those created in community hospital settings and stroke units tend to focus strongly on dysphagia identification and treatment [72]. It is possible that hospital settings without these traits may have even lower rates of dysphagia therapy communication within discharge summaries.

In conclusion, discharge summaries within this study frequently omitted critical dysphagia therapy recommendations made by hospital-based SLPs, even in populations at very high risk for aspiration. Future studies should focus on both improving the discharge communication of dysphagia therapy information and the impact this improved discharge communication has on patient outcomes, especially in vulnerable subacute care populations who rely strongly upon the systems that surround them.