In looking at the cost implications of our analysis, we must consider the payer mix when considering the implication of reducing ED and inpatient charges in such a drastic fashion, as insurance plans reimburse at variable rates. A 2016 Texas study found that for every $1.00 paid by Medicare to reimburse medical services, private insurance paid between$1.15 and $2.35, while Medicaid paid between $0.61 and $0.85.23 When looking at charges for services on the order of several million dollars, as in our study, the difference between reimbursement by private insurance and public insurance is enormous, also on the order of millions of dollars. In our study, the majority of patients had Medicaid insurance, which results in lower reimbursements to the hospital as compared to other insurance programs. While we were unable to perform a formal cost analysis of the charges and reimbursements to the hospital due to limitations in access to the data, the fact that our intervention reduced visits predominantly by patients with Medicaid insurance is not likely to be financially harmful to the hospital. Furthermore, in reducing charges by the patients in our program, our intervention was able to save significant monies for all insurance programs in our healthcare system,seedling grow rack which could be used for other health improvements and interventions, such as prevention and education. Finally, it is clear that our intervention – case management for ED frequent users – decreased ED visits, with the results evident from our study, as well as multiple previous studies cited above.
In our study, we noted a decrease in inpatient admissions, ED and inpatient LOS, charges, and the use of testing. The question arises as to whether case management reduces these metrics simply by keeping people out of the ED, or whether case management has some additional effect on utilization of services. In looking at Table 2, it becomes clear that ED visits decreased by 49%, with admissions and utilization of testing decreasing by about the same percentage, or slightly less. Continuing with Tables 3 and 4, LOS and charges decreased by less than 49%. This would suggest that the most effective aspect of ED case management for frequent users is the ability to decrease ED visits, with all other decreased metrics the result of the patient not being in the ED. Our study had several limitations. First, because we looked at ED and hospital visits at just one institution our study includes a relatively small number of patients.Thus, while we were able to significantly reduce cost, LOS, and utilization at our hospital, similar parameters may have increased at neighboring hospitals due to patients avoiding our institution. A study of the effect of ED case management on multiple hospitals within a geographic region would provide valuable information on this issue. Second, our study consisted of a retrospective chart review of a program in existence at our hospital, with no control group for comparison. While case management likely accounted for the significant changes in the parameters studied, it is possible that other factors, or simply regression towards the mean, accounted for part or all of our significant decreases. Another limitation was that we did not look at testing utilization over the long term, but rather only compared the year prior to the intervention to the year after the intervention.
For patients with recurrent complaints, physicians may not choose to perform imaging if imaging has recently been done. So, it is possible that robust imaging done on our patients in the year prior to enrollment decreased physician ordering of imaging studies in the year after enrollment. To be certain that our intervention decreased imaging study utilization, we would have needed to compare imaging in several years prior to enrollment to the year after enrollment. Finally, as previously mentioned we did not conduct a formal cost analysis of charges and reimbursements to our institution to determine the impact of the significant reduction in ED charges. While again we speculated that with the majority of enrolled patients having Medicaid, the reduced charges represented savings to the hospital, it is possible that the program may have reduced reimbursements to the hospital in an unfavorable way.Our multi-center prospective study examines and highlights the current practices and preferences for hand offs between EPs and PCPs and highlights quality gaps in the transition of the discharged ED patient back to the community. The study results suggest there is a discrepancy in provider expectations regarding best method of communication, and a disconnect between perception and reality of frequency of contact between ED and PCP providers. EPs preferred direct phone contact and communication synchronous to the encounter on patients needing urgent follow-up. In addition, EPs treated the communication of benign conditions differently that those with an urgent need. PCPs, on the other hand, preferred gathering information from the EMR and communication asynchronous to the encounter and wanted communication about non-urgent patients more often than EPs. EPs may prefer to communicate by telephone because perhaps they are not aware of the extent to which PCPs automatically receive updates through the EMR. Less than half of EPs perceived that PCPs receive an EMR notification, while a majority of PCPs reported receiving an alert through the EMR.
Since EMR notification was the preferred PCP method of communication, EPs might in future be more cognizant of the role of EMR notification to the PCP as a key component of transition of care for ED discharge. There are also existing,greenhouse growing racks under-used tools for communicating discharge information that are highly regarded and improve provider satisfaction.Limpahan et al. developed a set of best practices for patient discharge, including sending a summary to the PCP, performing medication reconciliation, and providing patient education.The authors suggest using the EMR as a potential avenue for automated inclusion of the described practices. Separate EMR systems have been identified as a challenge in the transition of care, while an interface for a shared EMR has been cited as a way to minimize transitions-of-care losses.Furthermore, the ability of EPs to provide an alert to the PCPs through flagging or email notification has been described as a potential tool for communication. In the present study providers reported setting and environmental constraints including a high patient volume, coordinating time to call, and communication during non-business hours. A standard EMR notification system may alleviate some of these constraints; however, EMR barriers to effective transitions were also noted, including lack of EMR access or shared EMR, uncertain receipt of information, and limited EMR literacy. Other logistical barriers to communication included inability to identify the PCP, difficulty getting in touch with the appropriate provider, and lack of resources. These are systems issues that could be addressed with increased emphasis on the ED-to-outpatient communication. Specific strategies might include readily available electronic documentation of a patient’s PCP, shared EMR access among hospitals and clinics, and professional coordinators to relay information during the discharge process. Healthcare providers believe that both technology and standardization should be the focus for future improvements in the transition of care.Shared EMR access and EMR notifications are potential areas for development. There are also new tools of clinical communication that may bridge the gap between the synchronous phone communication preferred by EPs and the asynchronous EMR communication preferred by PCPs.
Mobile health platforms that use HIPAA-compliant, secure text messaging can serve as an intermediate solution between phone message and EMR message, as these texts can satisfy the need for EPs to confirm delivery of an urgent message to a PCP, while allowing a small amount of asynchrony that does not disrupt the PCP’s workflow during a busy clinic day and is less intrusive than a phone call after hours. Further study is necessary to characterize the best structure and content of EMR notifications, in order to facilitate the transition of care from the ED to the outpatient setting. There are several limitations to this study. Most notably, the participants comprised a convenience sample of physicians from eight academic institutions. All community physicians worked at a community site affiliated with one of the primary academic sites. The present study lacks representation from community sites without academic affiliation, military, and rural institutions. Our responses may not reflect practice patterns in these settings. This study also lacks input from mid-level providers and residents who are also involved in the hand-off process. Interviews were performed in-person and therefore may have led to reporting bias on the part of the participant. Furthermore, this data is based on perception rather than objective measure of phone calls and EMR notifications, which is subject to recall bias. The qualitative questions served as a strategy to recruit more diverse responses. The process of coding synthesizes information, thus losing the context of specific statements in favor of categorizing data into themes. Finally, the majority of subjects in this study reported using EPIC EMR software. Other interfaces may allow for varying degrees of electronic communication between the ED and PCP, thus altering one’s perception of EMR utility. The concept of “direct to room” , also known as “immediate bedding,” has been reported in the literature as a mechanism to improve front-end emergency department processing.At one institution DTR was referred to as “closing” the waiting room, since patients were taken directly to a bed, when available, without undergoing formal triage and registration in the waiting room.Reducing wait times has been linked to patient perceptions of superior service and North well Health, Staten Island University Hospital, Department of Emergency Medicine, Staten Island, New York increased trust, especially in private hospitals.Although every ED may have individual front-end processes, most ED visits include patient presentation, registration, triage, bed assignment, and medical evaluation.Various models have been implemented in an attempt to reduce ED wait times and overall length of stay , from split flows to rapid triage.DTR uses the design of parallel processing, as opposed to serial processing, which allows patients to bypass many preliminary steps between arrival to the ED and placement in a bed. The goal is to decrease the backlog of waiting room patients waiting for less-critical tasks and allow registration, nursing evaluation, and medical provider evaluation to occur simultaneously at the bedside.More importantly, this facilitates an expedited clinician and patient interaction. The literature suggests that DTR can decrease waiting times, ED LOS, and left without being seen rates, while simultaneously improving patient satisfaction.Bertoty et. al. reported that the LOS for admitted patients decreased by 7.7%, and the LOS for discharged patients also decreased after DTR was implemented.Similarly, there was an improvement in patient satisfaction, which was hypothesized to occur since patients prefer to wait in a treatment area rather than a waiting room. Patients also perceived their treatment as beginning from the moment they were brought into the treatment area. At our institution, we implemented a DTR policy, which improved our front-end process dramatically. The Staten Island University Hospital ED has seen an improvement in metrics, similar to those cited in the literature, since implementing a DTR process. This includes decreased physician turn-around time, a decrease in LWBS, and a marked increase in patient satisfaction. Unfortunately, such improvements were accompanied by unforeseen consequences. In a traditional system, all patients undergo a formal triage process by a dedicated nurse, during which vital signs are obtained. In the DTR process, this step may be bypassed. Consequently, we noticed a delay from the time of presentation to the first recorded set of vital signs. In some circumstances, patients were unwittingly treated and released before obtaining a single set of vital signs. To address this issue, we developed a vital signs station within the waiting area. Our goal was to determine the feasibility and effectiveness of obtaining and recording vital signs within 10 minutes of every patient’s arrival to the ED after initiation of a DTR process.This retrospective, cohort study took place at a single, academic, tertiary-care, Level I trauma center with an annual census of approximately 94,000 visits. Inclusion criteria were all patients who entered the ED between the hours of 7 am to 11pm. We excluded from the study all patients who entered the ED between 11 pm and 7 am. due to inability to staff the vital signs station during these hours of the pilot phase of the program.