Improving Wait Time for Patients in a Pediatric Echocardiography Laboratory - a Quality Improvement Project

Supplemental Digital Content is available in the text.


INTRODUCTION
At our busy tertiary pediatric care center, prolonged wait time for echocardiograms (echo) in the outpatient pediatric cardiology clinic at the main campus was a source of patient and provider dissatisfaction with anecdotal reports of patients waiting for an hour or longer. The Institute of Medicine has defined timeliness, efficiency, and patient-centered care as 3 of 6 primary aims to help improve quality of health care in the United States. 1 Although the detrimental effect of prolonged patient wait times on the delivery of high-value health care is difficult to quantify, existing literature suggests that the impact extends beyond short-term patient satisfaction to the global perception of the quality of care received and the likelihood of continuing care with the same provider. [2][3][4][5] As the paradigm in health care delivery shifts toward value and quality, it has become increasingly important to improve efficiency and reduce waste.

SPECIFIC AIMS
To improve efficiency and patient and provider satisfaction in our outpatient pediatric cardiology clinic, we aimed to measure our baseline performance with regard to echo wait time (EWT) and then implement a Quality Improvement (QI) project to improve EWT. The S.M.A.R.T. (specific, measurable, achievable, realistic and timely) aim of our project was to decrease EWT for 90% of outpatient echos in the main cardiology clinic to < 20 minutes within 1 year of initiation of the project. 6

Context
We conducted this QI project in the outpatient echo laboratory at Children's Mercy Hospital, Kansas City. This high volume pediatric laboratory (annual volume = 17,000 echos) provides echo services at 10 locations across 2 states. Over 50% of all studies are performed at the main campus, where the outpatient pediatric cardiology facility comprises 10 clinic rooms and an on-site echo laboratory with 3 imaging rooms. Echocardiograms are performed on-demand for cardiology clinic patients while echos ordered by noncardiology providers including pediatric subspecialists and community pediatricians ("echoonly'') are performed on a scheduled basis. Outpatient throughput relies on both clinic and echo laboratory processes. Clinic patients are registered by a receptionist, moved from the waiting room by a care assistant (CA) who measures their height, weight, and vital signs; they then wait in a clinic room to meet with the provider. When an echo is required, a nurse activates the order; once it prints, it is walked to the echo laboratory and placed in a queue. The echo begins as soon as staff and space are available.

INTERVENTIONS
A QI team was formed comprising of cardiologists (A.P., T.S.), sonographer (A.W.), and QI mentor (J.M.). This initiative was approved by the Children's Mercy Hospital Institutional Review Board as a QI project that did not require documentation of informed consent. Key drivers of EWT were identified to focus interventions for improvement ( Fig. 1). Plan-do-study-act (PDSA) cycle interventions were then sequentially implemented with follow-up data collection to assess the effect of the interventions. The interventions were based on 3 key drivers: optimized space utilization, elimination of redundant steps and inefficiency, and shared ownership and improved communication among echo laboratory and clinic staff.

Repurposing and Maximizing Utilization of Space
Space was a significant limitation to improving patient throughput. Expansion of the clinic and echo laboratory footprint was not within the scope of this project, so alternative solutions were designed. We repurposed 1 patient room to an additional echo room to address the mismatch between the numbers of clinic and echo rooms. Nephrology and oncology echos were relocated to a repurposed imaging room in a separate (Dialysis) clinic. This redesign moved "echo only" patients off-site and allowed the 4 cardiology imaging rooms to be devoted to cardiology clinic patients. Additionally, compact imaging machines were used to perform echos in clinic rooms when the focus of the study was limited (eg, effusion check postcardiac surgery).
To maximize utilization of the 4 cardiology imaging rooms, the "Early Start" PDSA cycle addressed variable echo room utilization due to fluctuations in echo demand during clinic hours. Sonographers and space were available as early as 7 am, but there was poor utilization of the echo laboratory rooms at that time. Subsequently, the laboratory would receive clusters of orders, usually peaking between 9:30 and 10:00 am and resulting in a domino effect of causing delay. Utilization of all 3 imaging rooms by 8:00 am was established as a goal, resulting in adjustment of the "echo-only" schedule for earlier appointment times. We adjusted the cardiology clinic schedule so that it would start by 7:30 am, such that follow-up patients could have an echo performed early-even before provider evaluation.

Elimination of Redundancy and Inefficiency
Interventions implemented to improve efficiency were (1) elimination of redundant steps; (2) improvement in workflow by alternate electronic medical record (EMR) processes; and (3) recruitment of a dedicated CA for the echo laboratory. On analysis of the baseline data and through feedback from the sonographers, it became apparent that sonographers spent a significant proportion of their time performing tasks that could potentially be delegated. For example, they escorted patients between the clinic and the echo laboratory; they also cleaned rooms and equipment between patients and walked back and forth with echo billing sheets. To decrease the amount of foot traffic for paper billing, we implemented paperless billing to allow billing via the echo reporting system itself. We altered the automated printing process in the EMR so that activated echo orders would print in the sonographers' work area rather than in the clinic. The CA performed cleaning of rooms and equipment and escorted patients to and from their echo; as a result, sonographers were freed up to perform scans and create preliminary reports-tasks that were aligned with their expertise.

Shared Ownership and Improved Communication
Even though the physical space of the echo laboratory is in continuity with that of the clinic, the 2 systems operated in silos. To address suboptimal communication, we initiated huddles between the clinic and echo laboratory charge staff. Information shared during huddles included the anticipated numbers of clinic patients, clinic echos ordered, echo-only appointments and staffing, and resources including number of providers in clinic and sonographer availability. Another problem was the inability to track patients effectively as they moved through the process. A dry-erase board was being used; this was inconsistently updated by the sonographers and clinic nurses. We replaced this board with an EMR-based electronic whiteboard that allowed staff to track patients; it informed providers of the patient check-in time, patient location, and total time for clinic visit. The lead clinic nurse and echo laboratory technical director were tasked with the new responsibility of continuously assessing clinic and echo laboratory traffic throughout the day. Consistent communication and coordination allowed early identification of problems and proactive implementation of solutions.

STUDY OF THE INTERVENTIONS
We sequentially implemented interventions and monitored their effect on EWT by using Statistical Process Control charts to assess whether the interventions resulted in the desired change. We used established rules for differentiating special versus common cause variation for the charts. In addition, we compared the baseline and postintervention (after implementation of all the PDSA cycles) phases with regard to outcomes related to EWT using standard statistical tests.

MEASURES
We defined EWT as the time interval between activation of the echo order in the EMR and the first cine acquisition in the electronic echo image storage system; this was the primary outcome measure. To target patients at highest risk for prolonged EWT, we collected data on the busiest days of the clinic. A data collection form was posted in each echo room for sonographers to capture prospectively the following: time of echo order activation, patient arrival to echo room, acquisition of first echo clip, and patient departure from echo room.
In QI methodology, balancing measures are defined as unintended negative consequences of the project. Balancing measures that we considered included the possibility that emphasizing echo laboratory efficiency may adversely affect study quality. Although we did not grade echocardiograms for quality as part of this project, we monitored the time taken for image acquisition (calculated as the time taken from the first clip to the last clip on the study). We employed direct patient observation to shadow 15 patients from the time they checked into the clinic to the start of their echo. A flow-chart detailing the steps from patient registration until echo initiation allowed for the identification of potential causes for significant delay (Fig. 2).

Statistical Analysis
The primary outcome (EWT) was analyzed using statistical process control charts-X bar-S chart using Excel QI Macros. Special cause variation was defined as 8 data points above or below the baseline mean. Descriptive statistics are reported as means ± SD for continuous variables and frequency with percentage for categorical variables. Differences in outcomes between intervention groups were assessed using 2-sided independent t-tests for continuous variables and Chi-square or Fisher's exact test for categorical variables, as appropriate based on cell size. Time spent in echo was highly skewed and therefore reported as median and interquartile range (IQR) and tested using Wilcoxon rank-sum test. All statistical tests were 2-sided and conducted at the alpha = 0.05 level. Statistical analysis was done using the SAS software v 9.4 (SAS Institute Inc., Cary, N.C.) and R (R Core Team, 2015: R Foundation for Statistical Computing, Vienna, Austria).

RESULTS
Between January 27, 2014, and August 27, 2015, we tracked 840 patients for EWT. To assess if the improvement was sustained, sampling was again performed from January 22, 2016, to March 25, 2016. We divided the project into 7 phases with phase 1 being the preintervention phase, phase 7 being the postintervention phase, and the phases in between corresponding to the interventions that were implemented; these are delineated in Table 1.  (red lines) indicate a more predictable process postintervention. A major downward shift in the weekly running mean occurred in late December 2014 with the repurposing of a clinic room as an additional echo laboratory, which also coincided with the functioning of the lead sonographer as echo laboratory traffic manager and effectively allocating resources. Figure 4 shows the S-chart of the weekly EWT; each point represents the SD for EWT for the week, and the center line (solid yellow line) represents the mean of the SDs. Similar to the X-bar chart, a downward shift is seen after implementation of the third PDSA cycle. There is a narrowing of the overall confidence limits; however, there is still variation from patient to patient that is not yet predictable within the control limits. Table 2 depicts the characteristics of EWT at baseline and postintervention. Compared with baseline, the postintervention mean and maximal wait times decreased significantly (22.5 ± 17.5 minutes versus 15.3 ± 7.8 minutes, P < 0.0001, and 83 minutes versus 47 minutes, P < 0.001, respectively). Postintervention, there was a significant increase in patients waiting < 20 minutes (81% versus 62%, P < 0.0001) and < 30 minutes (97% versus 76%, P < 0.0001), respectively.

DISCUSSION
In this project, we demonstrate that process observation can be used to drive interventions that resulted in a significant and sustained improvement in wait times for echocardiograms in a busy outpatient echo laboratory setting. To our knowledge, this is the first report of a QI project to improve echo wait times in the outpatient cardiology setting.

Choice and Design of this Project
We recognized echo wait time as a significant problem that was affecting satisfaction for both our patients and our providers. To have a specific, measurable, and well-defined aim, we defined EWT as the time from activation of the order in the EMR until the first echo clip  was recorded, understanding that this also included the time taken to walk the patient to the echo laboratory and prepare them for their echo. Limiting the definition of EWT in this manner meant that EWT could be tracked retrospectively if the sonographers did not record times on their time sheet. To ensure a manageable burden of data collection, we sampled only those days that were busy for ~2 mornings a week. This strategy suggests that overall performance may be better than what was measured in this project; nevertheless, we were able to study better the days that were most adversely affected and gain insight into the nature and causes of the delays and also ensure that data collection was complete and accurate.

Lessons Learned
We realized that process observation is crucial to identify areas in need of improvement. Along with initial discussions with clinic and echo laboratory personnel, process observation identified several time-consuming steps and processes that could easily be eliminated. A prioritization matrix can help select interventions and the sequence in which to implement them (Supplementary figure 1,

available as
Supplemental Digital Content at http://links.lww.com/ PQ9/A29). The scale of the desirable effect of the intervention has to be balanced with the ease of implementation-the ideal intervention being high impact and requiring low to medium effort. Interventions such as paperless processes, elimination of back and forth walking with billing sheets and orders, and alteration of EMR workflow to print orders in the sonographer workspace were easy to implement and represented "low hanging fruit." We also considered changing our echo laboratory flow from as-needed to scheduled echos for clinic patients. However, this was determined to be too drastic a change with potential undesired effects on patient experience. Similarly, implementing a system-wide change of allowing CA's to have EMR privileges to activate orders was considered impractical and was therefore not pursued. Recruiting a CA dedicated to the echo laboratory required buy-in from the leadership due to the financial burden posed, but was pursued because it projected to have a significant impact on workflow. It was important to achieve buy-in from key stakeholders (sonographers, clinic nurses, and physicians, Heart Center leadership); we appointed champions for the project in each area and held monthly meetings to update all involved. There is a narrowing of the overall confidence limits; however, there is still variation from patient to patient that is not yet predictable within the control limits. Out-of-the-box thinking helped us work around problems such as lack of space. Outsourcing echo-only studies to the Dialysis clinic improved patient flow for our renal patients, who previously had to walk from that clinic to the echo laboratory. Since then, our group has adopted this model of care for other specialty clinics such as the muscular dystrophy clinic and the cardio-oncology clinic. The use of a compact machine in clinic rooms to do quick follow-up echocardiograms allowed us to achieve additional efficiency. The designation of our lead sonographer as a "traffic cop" allowed for the continual assessment and redistribution of the sonographer pool between inpatient and outpatient areas to meet clinical needs.
QI projects require that the team identify and track balancing measures. There was a significant increase in the median time taken for completion of an echo from pre-to postintervention; this could be considered as indirect evidence that quality was not compromised in the interest of speed. Changes in the responsibilities of echo and clinic staff could also adversely affect job satisfaction, and optimization of clinic processes for echo throughput could adversely affect other clinical areas such as exercise or ECG laboratory workflow. Although we did not track these parameters specifically, there were no specific instances of provider/staff dissatisfaction.
As might be expected in any real-life project, we faced hurdles; a severe sonographer staffing shortage required a temporary halt to the study between August and December 2014. Not all providers were eager to embrace changes; regular and structured updates helped in this regard. Overall, this project shifted the culture among the clinic and echo laboratory personnel from a sense of helplessness to that of awareness and responsiveness with regard to wait time for patients.
Ensuring high quality in echocardiography is recognized as an important goal of individual echo laboratories; to that end, guidelines have been published by several organizations including the American Society of Echocardiography, American College of Cardiology, Inter-societal Accreditation Commission and the European Society of Echocardiography. [7][8][9] Most of the published QI work in echocardiography has focused on improving interobserver and interacquisition variability in quantitative data or compliance with established guidelines. [10][11][12][13][14] We have previously shown that implementation of a QI program results in significant improvement in image quality and completeness of echocardiograms in postoperative congenital heart disease patients. 15 In this QI study, we focused on improving timeliness, which is less studied, but integral to patient care.

Limitations
The findings of this QI project may not be generalizable across echo labs of varying structure and volume; nevertheless, we have outlined the methodology that we have used and the rationale for implementing specific interventions, which could be of use to others designing such a project. This project only tracked the EWT before commencement of the echo, we did not track any wait time that patients may have experienced while waiting for the finalized report after the echo was completed. We did not measure patient satisfaction directly because we did not have a system where this could be measured independent of the total clinic experience; therefore, it is not known if improving EWT had the intended effect of improving patient satisfaction. We also did not conduct a formal survey of the providers to assess satisfaction with the new state of patient throughput; we, however, received positive feedback informally from presentations at faculty meetings.

CONCLUSIONS
We were able to significantly decrease echocardiography wait times in our ambulatory clinic by using QI methods. In an era when physicians are required to participate in QI, it is important for individual health care teams to devise and implement QI processes that are meaningful to their system.