Across the Continuum: How Inpatient Palliative Care Consultations Are Reported in Hospital Discharge Summaries

Introduction

Palliative care (PC) is an interdisciplinary approach to the management of serious illness that focuses on decision making, pain and symptom management, and navigation of a complex healthcare system. ¹ The number of patients who are in need of PC services worldwide has grown to an estimated 40 million but only about 14% actually receive it. ² Inpatient PC consultations focus on providing information about prognosis and establishing goals of care for patients with serious and often terminal illnesses. The identified goals determine the plan of care and must be communicated to future providers and sites of care across the continuum to have more effective transitions of care. Because discharge summaries (DSs) are an essential tool for communication in a transition of care, it is critical that the information discussed in PC consultation be documented in the DS.

From the system perspective, a “transition of care” is a set of actions designed to ensure the coordination and continuity of healthcare a patient transfers between different locations or levels of care. ³ Mismanaged handoffs and poorly executed transitions are a cause of avoidable but costly errors, breakdowns in care plans, and preventable rehospitalizations. ⁴ Lack of notation in the DS regarding the outcomes of PC consultation could create an incomplete depiction of the patient's condition, illness understanding, and goals. Efforts to improve transitions could save money by preventing expensive and avoidable complications that result in rehospitalizations or extended postacute care stays. ³ Accurate communication between the primary team and PC and appropriate documentation of the palliative plan in the DS is essential for an effective transition in care.

Documentation of family meetings is another element that contributes to the representation of a PC patient. Family meetings provide emotional support to the family, elicit patient's goals and values, and build trust. ⁵ It is common that decisions central to the patient's goals and plan are made in these settings. Without record of a family meeting in the DS, future providers may not know whether the family's needs have been addressed or what discussions are needed. Family meeting status is an important aspect that should be in the DS.

Materials and Methods

This retrospective review of 961 charts included all patients 18 years or older who received an inpatient PC consultation at the University of Kansas Hospital from July 2011 to May 2015, who were discharged alive and not referred to hospice. Hospice patients were excluded because their goals of care were established before discharge. The charts were extracted by the Hospital Information Technology department based on the criteria. A list of code words and their related phrases, developed by an expert panel of geriatricians and PC physicians in 2013, was used as markers of communication and information from a PC consult in evaluation of completeness of the DS. Before evaluation of the charts, the same panel of physicians reviewed the first six DSs and validated the code words. ⁶ The code words and their related phrases were broken into five main categories: PC, symptom management, hospice and palliative home health, decision making, and plan of care (Table 1). The data were collected during two different time frames. From July 2011 to June 2013, 366 charts were evaluated for code words and demographics. From July 2013 to May 2015, 595 charts were evaluated for the same criteria from the first time frame, along with reviewing documented communication between the primary team and PC and also documented family meetings in the PC consult and notes. The primary investigator reviewed each DS for the initial code words and related phrases and recorded their frequency of usage. ⁶ These were then sorted into the five main categories for analysis. A second author reviewed 22% of the charts to ensure reliability of the PI reviewer's data collection; discrepancies occurred in 2% of the charts between the PI and the second author's data collection. Logistic regression models were used to examine the associations between explanatory variables and the outcome of code words in DSs. Limitations include an one-site study and investigators were unable to evaluate verbal communication between providers.

Results

Characteristics of patients in 961 charts reviewed showed 49% males, ages 18–102 years with a mean age of 65.2 years (Table 2). The average length of stay was 16.7 (standard deviation [SD] = 20.9) days and average time from PC consultation to discharge was 10.1 (SD = 17.0) days. Among these patients, 53.7% were “Full Code” (FC), 39.5% were “Do Not Attempt Resuscitation” (DNAR), and 6.8% were “Limited Attempt at Resuscitation” (LAR). Discharging services included internal medicine (77.1%), surgery (19.7%), and burn (3.2%). About half (49.7%) of patients were discharged to home, half (24.2% of total) of whom were sent home with home health; the remainder were discharged to a skilled nursing facility (28.8%), long-term acute care hospitals (14.8%), outside hospital (2.2%), or inpatient rehabilitation facility (4.3%). There were no code words included in 22.8% of the DSs. PC was mentioned in 63.3% of DSs and was the only code word in 5.3%. Symptom management was discussed in 37.0%, plans of care in 36.4%, decision making and capacity in 34.9%, and hospice and palliative home health in 17.2% (Fig. 1). Most records (92.3%) of PC consults and notes documented communication between the primary team and PC and only 38.5% documented a family meeting. Among the 38.5% of records with documentation of a family meeting, only 12% mentioned it in the DSs. Only days from PC consultation to discharge (odds ratio [OR] = 0.975, 95% confidence interval [CI] 0.964–0.986, p value <0.001) and code status (DNAR/LAR vs. FC OR = 1.523, 95% CI 1.112–2.086, p value = 0.009) were found to be significantly associated with the outcome of code words in their DSs.

OPEN IN VIEWER

FIG. 1.

The percentage of code words and their related phrases found in discharge summaries.

Discussion

The primary goal of this study was to identify how the outcomes of inpatient PC consultation are documented in the DSs and ways to improve the documentation system for more effective transitions of care. Improving transitions could increase coordination of providers across the continuum, increase patient centeredness of care, improve concordance between patient preferences and plan, and reduce rehospitalization rates and healthcare costs. To improve the continuity of care, the outcomes of consultations should be communicated to all providers. When there was documentation of primary and PC team communication, the DSs tended to have more code words. This demonstrates that communication between teams is essential for more complete documentation in the DS. If family meeting status was documented, more code words were detected. A family meeting is an important tool for providers to facilitate communication for people with advanced illness. ⁷ Studies analyzing family meetings have shown to reduce the length of hospitalization, which reduces costs. ⁸ Our results suggest that it may be beneficial to include a family meeting section in all DSs, to spur the writer to include such relevant information.

The longer the patient remained in the hospital after the consultation, the less likely there were code words in the DS. Our study shows the potential for a positive impact on patient care if the electronic medical record included a section to list all consultations or if it automatically listed all consultations during the hospitalization in the DS. This would increase the ability for the provider to quickly read consultations and to generate a more complete DS.

Patients with a limitation in their desire for resuscitative efforts, as reflected in a DNAR or LAR code status, were significantly more likely to have DSs containing code words. This reflects that patients with code status limitations in their chart had more communication of the PC consultation reflected in the DSs. This may be because the code status is easily seen in the chart.

Secondary analysis identified that males had more code words than females. Patients with an older age had more code words. Patients who were discharged to home or home with home healthcare tended to have fewer code words.

More than one out of five DSs had no discussion of PC. DSs are the traditional means of provider communication and more guidelines are needed to ensure that the patient's full visit history is reflected in it. We propose adding a goals-of-care section, family meeting section, and a list of all consultations to all standard DSs to improve communication between sites of care.

Conclusion

More than one in five DSs after a hospitalization that included a PC consultation had no mention of any code words and more than one in three DSs lacked mention of PC. DSs should reflect the interventions and outcomes of PC consultation to effectively communicate to all receiving providers, facilitate a better transition of care, and meet the needs of patients. We propose documenting all consultations, adding a goals-of-care section, and adding a family meeting status section to all standard DSs. We hypothesize that these changes will improve documentation and communication and then future studies can compare these outcomes. With these changes, providers would be better equipped to improve the quality of care for all people involved in the patient's care. Further interventions to improve quality of provider communication are needed.