The hospital discharge process: Call for technology’s help

While being discharged from the hospital even after a minor procedure is not simple, the process for a patient with comorbidities after a prolonged stay is daunting.

Physicians from multiple specialties, various nonphysician providers, the social worker, and the case manager all address different discharge-related issues. It is frustrating for both a provider and patient to experience the "I really can’t answer that question" moment. Lack of interdisciplinary communication may lead to medical errors and either premature or delayed discharges.

The date of discharge is estimated soon after admission. Some hospitals have a focus on the clock when planning discharges. If planning occurs too early, it does not account for changes in patient needs and wrong instructions might be given. Transportation and home-aide needs are time sensitive.

In contrast, some planning does need to be considered early in the admission when discharge to a non-acute care facility is obvious due to the diagnosis and/or social situation of the patient.

One study from the Brigham and Women’s Hospital identified seven clinical factors predicting hospital readmission: a hemoglobin less than 12 g/dL on discharge, discharge from an oncology service, low serum sodium level on discharge, a procedure (via ICD-9 standards) during admission, nonelective admission, length of stay greater than 4 days, and number of admissions during the previous year (JAMA Intern. Med. 2013;173:632-8).

Another study examined many predictive models found in the literature.

The researchers found that "of 7,843 citations reviewed, 30 studies of 26 unique models met the inclusion criteria. The most common outcome used was 30-day readmission; only 1 model specifically addressed preventable readmissions. Fourteen models that relied on retrospective administrative data could be potentially used to risk-adjust readmission rates for hospital comparison; of these, 9 were tested in large U.S. populations and had poor discriminative ability. ... Seven models could potentially be used to identify high-risk patients for intervention early during a hospitalization, ... and 5 could be used at hospital discharge" (JAMA 2011;306:1688-98).

The authors concluded that most prediction models perform poorly or require improvement. Perhaps one reason for this result lies in the fact that these models traditionally are either clinical or administrative. I believe a better approach is to combine administrative and clinical predictive models. Better analytics programs applied real-time in the electronic health record (EHR) will facilitate integration of these perspectives.

The topic of transitional care has received attention because a poor discharge process results in higher readmission rates, a new benchmark focus of Medicare (Am. J. Nurs. 2008;108:58-63). Hospitals might be very good at meeting regulatory requirements, but the patient’s understanding of diagnoses and instructions is often unclear. Though required by regulations, the caregiver may not even be included in the process. Technology can help in this situation. Some of possibilities mentioned below might not be available in the context described.

• Durable equipment needs. The care coordinator is generally the point person regarding the patient’s durable equipment needs upon discharge. Ordering the equipment (specifications as well as date, time, and place of delivery) might be the job of someone else, such as a therapist or physician. Digital tools can expedite equipment procurement. Analytics from the EHR (mining diagnoses, equipment in use at the end of the hospitalization, expected place of transition, etc.) might determine the individual’s ambulation, oxygen, bed, or other equipment requirements. This can act as a preliminary checklist for the coordinator, doing away with the need to personally go through the EHR or surveying providers. A digital ordering program can directly interact with the distributor to check product availability and verify delivery. Another useful tool would aggregate equipment distributors, which are stratified according to certification (Medicare bidding approval status), lowest price, and best-rated service.

• Visiting nurses. Often the home-needs assessment for visiting nurses is done once the patient is discharged. This can be expedited with the help of a caregiver, with the assessment completed in the hospital. Consider a tool into which the physician’s orders or recommendations for home nursing are placed and shared with the visiting nurse entity, the patient, and the caregiver. It would include the nursing assessment as well as a video of the home environment (a factor in the assessment itself). This would obviate the need for a dedicated assessment visit. Visiting nurses themselves should be equipped with mobile technology to document their time schedule for billing, to record interventions, and to record and transmit vital signs (measured via digital remote monitors) and orders; the technology also should contain a digital messaging program.

• Scheduling of outpatient provider appointments. Evidence suggests that in a general medical population, early follow-up appointments do not affect readmission rates (Arch. Intern. Med. 2010;170:955-60). However, some patients, including those with heart failure have been shown to benefit from early follow-up (JAMA 2010;303:1716-22). The success of a growing number of commercially available mobile apps intended to streamline scheduling of physician appointments is testimony to this need in the nonacute setting. Patient portal use is a requirement of Meaningful Use Stage 2. One way of encouraging patient participation in portal use would be activating it by utilizing a discharge planning scheduling application of the portal at the time of discharge. This also fits into an overall strategy of point of engagement implementation of technology.

These are only a few highlights of the complexity of the discharge process. All physicians have dealt with the many questions, complications, and frustration experienced by patients after discharge. A failed process creates unnecessary work, expense, and bad outcomes.

To many physicians, digital health technology is represented by the EHR in its present form, which is not what the doctor ordered. It is not intuitive, it is cumbersome, and it encourages impersonal encounters with patients. I will explore in future posts how digital technologies other than the EHR will change medicine in ways that physicians will appreciate.

Dr. Scher, a practicing cardiac electrophysiologist in Lancaster, Pa., is director at DLS Healthcare Consulting, advising technology companies and health care enterprises on development and adoption of mobile health technologies.

While being discharged from the hospital even after a minor procedure is not simple, the process for a patient with comorbidities after a prolonged stay is daunting.

Physicians from multiple specialties, various nonphysician providers, the social worker, and the case manager all address different discharge-related issues. It is frustrating for both a provider and patient to experience the "I really can’t answer that question" moment. Lack of interdisciplinary communication may lead to medical errors and either premature or delayed discharges.

The date of discharge is estimated soon after admission. Some hospitals have a focus on the clock when planning discharges. If planning occurs too early, it does not account for changes in patient needs and wrong instructions might be given. Transportation and home-aide needs are time sensitive.

In contrast, some planning does need to be considered early in the admission when discharge to a non-acute care facility is obvious due to the diagnosis and/or social situation of the patient.

One study from the Brigham and Women’s Hospital identified seven clinical factors predicting hospital readmission: a hemoglobin less than 12 g/dL on discharge, discharge from an oncology service, low serum sodium level on discharge, a procedure (via ICD-9 standards) during admission, nonelective admission, length of stay greater than 4 days, and number of admissions during the previous year (JAMA Intern. Med. 2013;173:632-8).

Another study examined many predictive models found in the literature.

The researchers found that "of 7,843 citations reviewed, 30 studies of 26 unique models met the inclusion criteria. The most common outcome used was 30-day readmission; only 1 model specifically addressed preventable readmissions. Fourteen models that relied on retrospective administrative data could be potentially used to risk-adjust readmission rates for hospital comparison; of these, 9 were tested in large U.S. populations and had poor discriminative ability. ... Seven models could potentially be used to identify high-risk patients for intervention early during a hospitalization, ... and 5 could be used at hospital discharge" (JAMA 2011;306:1688-98).

The authors concluded that most prediction models perform poorly or require improvement. Perhaps one reason for this result lies in the fact that these models traditionally are either clinical or administrative. I believe a better approach is to combine administrative and clinical predictive models. Better analytics programs applied real-time in the electronic health record (EHR) will facilitate integration of these perspectives.

The topic of transitional care has received attention because a poor discharge process results in higher readmission rates, a new benchmark focus of Medicare (Am. J. Nurs. 2008;108:58-63). Hospitals might be very good at meeting regulatory requirements, but the patient’s understanding of diagnoses and instructions is often unclear. Though required by regulations, the caregiver may not even be included in the process. Technology can help in this situation. Some of possibilities mentioned below might not be available in the context described.

• Durable equipment needs. The care coordinator is generally the point person regarding the patient’s durable equipment needs upon discharge. Ordering the equipment (specifications as well as date, time, and place of delivery) might be the job of someone else, such as a therapist or physician. Digital tools can expedite equipment procurement. Analytics from the EHR (mining diagnoses, equipment in use at the end of the hospitalization, expected place of transition, etc.) might determine the individual’s ambulation, oxygen, bed, or other equipment requirements. This can act as a preliminary checklist for the coordinator, doing away with the need to personally go through the EHR or surveying providers. A digital ordering program can directly interact with the distributor to check product availability and verify delivery. Another useful tool would aggregate equipment distributors, which are stratified according to certification (Medicare bidding approval status), lowest price, and best-rated service.

• Visiting nurses. Often the home-needs assessment for visiting nurses is done once the patient is discharged. This can be expedited with the help of a caregiver, with the assessment completed in the hospital. Consider a tool into which the physician’s orders or recommendations for home nursing are placed and shared with the visiting nurse entity, the patient, and the caregiver. It would include the nursing assessment as well as a video of the home environment (a factor in the assessment itself). This would obviate the need for a dedicated assessment visit. Visiting nurses themselves should be equipped with mobile technology to document their time schedule for billing, to record interventions, and to record and transmit vital signs (measured via digital remote monitors) and orders; the technology also should contain a digital messaging program.

• Scheduling of outpatient provider appointments. Evidence suggests that in a general medical population, early follow-up appointments do not affect readmission rates (Arch. Intern. Med. 2010;170:955-60). However, some patients, including those with heart failure have been shown to benefit from early follow-up (JAMA 2010;303:1716-22). The success of a growing number of commercially available mobile apps intended to streamline scheduling of physician appointments is testimony to this need in the nonacute setting. Patient portal use is a requirement of Meaningful Use Stage 2. One way of encouraging patient participation in portal use would be activating it by utilizing a discharge planning scheduling application of the portal at the time of discharge. This also fits into an overall strategy of point of engagement implementation of technology.

These are only a few highlights of the complexity of the discharge process. All physicians have dealt with the many questions, complications, and frustration experienced by patients after discharge. A failed process creates unnecessary work, expense, and bad outcomes.

To many physicians, digital health technology is represented by the EHR in its present form, which is not what the doctor ordered. It is not intuitive, it is cumbersome, and it encourages impersonal encounters with patients. I will explore in future posts how digital technologies other than the EHR will change medicine in ways that physicians will appreciate.

Dr. Scher, a practicing cardiac electrophysiologist in Lancaster, Pa., is director at DLS Healthcare Consulting, advising technology companies and health care enterprises on development and adoption of mobile health technologies.

While being discharged from the hospital even after a minor procedure is not simple, the process for a patient with comorbidities after a prolonged stay is daunting.

Physicians from multiple specialties, various nonphysician providers, the social worker, and the case manager all address different discharge-related issues. It is frustrating for both a provider and patient to experience the "I really can’t answer that question" moment. Lack of interdisciplinary communication may lead to medical errors and either premature or delayed discharges.

The date of discharge is estimated soon after admission. Some hospitals have a focus on the clock when planning discharges. If planning occurs too early, it does not account for changes in patient needs and wrong instructions might be given. Transportation and home-aide needs are time sensitive.

In contrast, some planning does need to be considered early in the admission when discharge to a non-acute care facility is obvious due to the diagnosis and/or social situation of the patient.

One study from the Brigham and Women’s Hospital identified seven clinical factors predicting hospital readmission: a hemoglobin less than 12 g/dL on discharge, discharge from an oncology service, low serum sodium level on discharge, a procedure (via ICD-9 standards) during admission, nonelective admission, length of stay greater than 4 days, and number of admissions during the previous year (JAMA Intern. Med. 2013;173:632-8).

Another study examined many predictive models found in the literature.

The researchers found that "of 7,843 citations reviewed, 30 studies of 26 unique models met the inclusion criteria. The most common outcome used was 30-day readmission; only 1 model specifically addressed preventable readmissions. Fourteen models that relied on retrospective administrative data could be potentially used to risk-adjust readmission rates for hospital comparison; of these, 9 were tested in large U.S. populations and had poor discriminative ability. ... Seven models could potentially be used to identify high-risk patients for intervention early during a hospitalization, ... and 5 could be used at hospital discharge" (JAMA 2011;306:1688-98).

The authors concluded that most prediction models perform poorly or require improvement. Perhaps one reason for this result lies in the fact that these models traditionally are either clinical or administrative. I believe a better approach is to combine administrative and clinical predictive models. Better analytics programs applied real-time in the electronic health record (EHR) will facilitate integration of these perspectives.

The topic of transitional care has received attention because a poor discharge process results in higher readmission rates, a new benchmark focus of Medicare (Am. J. Nurs. 2008;108:58-63). Hospitals might be very good at meeting regulatory requirements, but the patient’s understanding of diagnoses and instructions is often unclear. Though required by regulations, the caregiver may not even be included in the process. Technology can help in this situation. Some of possibilities mentioned below might not be available in the context described.

• Durable equipment needs. The care coordinator is generally the point person regarding the patient’s durable equipment needs upon discharge. Ordering the equipment (specifications as well as date, time, and place of delivery) might be the job of someone else, such as a therapist or physician. Digital tools can expedite equipment procurement. Analytics from the EHR (mining diagnoses, equipment in use at the end of the hospitalization, expected place of transition, etc.) might determine the individual’s ambulation, oxygen, bed, or other equipment requirements. This can act as a preliminary checklist for the coordinator, doing away with the need to personally go through the EHR or surveying providers. A digital ordering program can directly interact with the distributor to check product availability and verify delivery. Another useful tool would aggregate equipment distributors, which are stratified according to certification (Medicare bidding approval status), lowest price, and best-rated service.

• Visiting nurses. Often the home-needs assessment for visiting nurses is done once the patient is discharged. This can be expedited with the help of a caregiver, with the assessment completed in the hospital. Consider a tool into which the physician’s orders or recommendations for home nursing are placed and shared with the visiting nurse entity, the patient, and the caregiver. It would include the nursing assessment as well as a video of the home environment (a factor in the assessment itself). This would obviate the need for a dedicated assessment visit. Visiting nurses themselves should be equipped with mobile technology to document their time schedule for billing, to record interventions, and to record and transmit vital signs (measured via digital remote monitors) and orders; the technology also should contain a digital messaging program.

• Scheduling of outpatient provider appointments. Evidence suggests that in a general medical population, early follow-up appointments do not affect readmission rates (Arch. Intern. Med. 2010;170:955-60). However, some patients, including those with heart failure have been shown to benefit from early follow-up (JAMA 2010;303:1716-22). The success of a growing number of commercially available mobile apps intended to streamline scheduling of physician appointments is testimony to this need in the nonacute setting. Patient portal use is a requirement of Meaningful Use Stage 2. One way of encouraging patient participation in portal use would be activating it by utilizing a discharge planning scheduling application of the portal at the time of discharge. This also fits into an overall strategy of point of engagement implementation of technology.

These are only a few highlights of the complexity of the discharge process. All physicians have dealt with the many questions, complications, and frustration experienced by patients after discharge. A failed process creates unnecessary work, expense, and bad outcomes.

To many physicians, digital health technology is represented by the EHR in its present form, which is not what the doctor ordered. It is not intuitive, it is cumbersome, and it encourages impersonal encounters with patients. I will explore in future posts how digital technologies other than the EHR will change medicine in ways that physicians will appreciate.

Dr. Scher, a practicing cardiac electrophysiologist in Lancaster, Pa., is director at DLS Healthcare Consulting, advising technology companies and health care enterprises on development and adoption of mobile health technologies.