Decreasing waiting time for new patients at a community pain clinic

Abstract

Long waiting time to access pain medicine clinics poses a significant mental, physical, and socioeconomic burden on patients with chronic pain. This project aimed to develop interventions to reduce the waiting time for new referrals. We used the define, measure, analyze, improve, control (DMAIC) method. Clinic data were analyzed over a 6-month period. Pilot interventions were then implemented in one provider’s clinic over a 3-month period. Outcome measures included the number of new patients seen, number of “no shows,” and number of patients on the waitlist. Late cancellation and no shows were the main causes of the clinic lost time. Interventions to reduce unutilized clinic time were implemented, including making appointment reminder calls, identifying cancellations in advance, and adding slots on the provider’s template to account for cancellations and no shows. These interventions resulted in a 16% decrease in no shows, a 60% increase in new patients seen, and a significant 47% reduction in the number of patients on the entire clinic waitlist. These findings suggest that simple procedures and changes in the clinic identified via a quality improvement process can significantly improve clinic time utilization.

Chronic pain is a debilitating condition that has a detrimental impact on patients’ health status, function, and quality of life and poses a considerable socioeconomic burden.^1,2 It is a complex and multidimensional problem that requires a multidisciplinary approach, which often cannot be provided by primary care providers alone. However, access to pain clinics is challenging, and a long appointment waiting time for new referrals is common.^3–5 Long appointment waiting times (several months) for pain clinic care may have deleterious consequences, including patient dissatisfaction, delay in diagnosis and treatment, prolonged sick leave, reduced functionality, and increased anxiety and depression.^6–8 On the other hand, patients who receive treatment in <2 months demonstrate greater improvements in quality of life and pain intensity compared to those who wait longer for their initial visit.^9,10 The aim of this project was to identify the potential causes of our clinic’s long waiting time for the initial appointment and to test interventions that can be used to decrease that waiting time.

METHODS

Our clinic receives internal and external referrals from a variety of healthcare providers in a large metropolitan area. Referrals are triaged by a trained registered nurse and then by one of the clinic pain medicine physicians before approval and scheduling of a new patient appointment. The approved referral is usually added to a new patient waitlist unless it is determined that the referral is urgent and needs an earlier appointment (through overbooking or use of a designated procedure slot) (Figure 1). The clinic is staffed by two pain medicine specialists, an anesthesia resident, and a pain medicine fellow. Each physician’s template is divided into a half day of seeing clinic patients (new patients or follow-ups) and a half day of procedures. For years, our clinic has had a waiting time of 3 to 6 months for new appointments. This problem led to complaints from patients and referring physicians. The waiting time was around 6 months before the beginning of our project.

Figure 1.

The process map of our clinic scheduling new patients.

We utilized the define, measure, analyze, improve, control (DMAIC) methodology for quality improvement design.¹¹ In accordance with our institutional review board, this quality improvement project, involving only deidentified information, was exempt from review.

Clinic data were obtained before and after the interventions through the institution’s electronic medical record. Collected data included number of approved referrals and available slots for new patients every month, details of providers’ templates, daily nonutilized clinic time, and patient disposition after the pain clinic visits. After the interventions, data were collected for number of new patients seen in the provider’s clinic, number of “no shows” in the provider’s clinic, and number of patients on the entire clinic waitlist.

RESULTS

As shown in Figure 2, there were more available slots for new patients than new approved referrals in the same month. Figure 3 shows the distribution of the pain clinic unutilized slots over the 6-month analysis period (a median of 10 slots). With input from the clinic pain medicine physicians, residents, fellows, and nurses, the causes for the long waitlist were analyzed, and lost clinic time was a major cause (Figure 4). To improve clinic time utilization, we focused on the three main modifiable causes for the clinic lost time: no shows, late cancellations, and the use of clinic slots for unnecessary follow-up visits.

Figure 2.

Comparison of approved new referrals to the available slots for new patients in our clinic.

Figure 3.

Box and whisker plots of the unutilized slots in our clinic during the data collection period from June 2017 to November 2017.

Figure 4.

Fishbone (cause-and-effect) diagram showing the possible causes of the long waiting time for initial appointments in our clinic.

Pilot interventions were then implemented in response to the analysis in one provider’s clinic (20% of the entire clinic census) from March 1, 2018, to May 31, 2018. These interventions included calling clinic patients a week in advance to remind them of their appointments; furthermore, the clinic schedule was reviewed 3 to 4 days in advance to fill cancellations with patients from the clinic waitlist. Also, two clinic slots were added for new patients on the provider’s clinic template to account for last-minute cancellations and no shows. To further improve the availability of the clinic slots for new patients, appropriate follow-ups were replaced with discharges or changed to as-needed follow-ups.

The above interventions resulted in an overall 16% decrease in no shows and an increase of 60% of new patients seen in the provider’s clinic (Figure 5). The number of the entire clinic waitlist decreased 47%, from 183 before the interventions to 87 after the interventions. After the 3-month intervention period, the entire clinic adapted some of the above interventions, which helped to further decrease the new patient waiting time. Other factors also helped to decrease that time, such as the adoption of a new triage system and the coronavirus pandemic. Currently our waiting time for new patients is <1 month.

Figure 5.

Run chart showing the improvement of the daily census of new patients seen in the study pilot clinic. The improvement phase started at week 12.

DISCUSSION

Overall, our project results revealed that no shows, late cancellations, and unnecessary follow-ups by far were the main causes of the unavailability of clinic slots for the new approved referrals. Implementation of simple changes identified via the DMAIC quality improvement process led to a dramatic improvement in the volume of new patients seen in the clinic and a decrease in the clinic no shows—and subsequently a significant decrease in our entire clinic waitlist for new patients.

The adoption of a phone call reminder for the patients a week before their appointments reminded the patients of their scheduled appointment, as considerable time had lapsed since the initial referral was placed. It also identified those who would not be able to make it or those with appointments that were no longer needed, and did so early enough to make changes to the providers’ schedule without causing disruption to the clinic workflow. As a result, it reduced late cancellations and no shows, major contributors to nonutilized clinic time. A systematic review by Ansell et al on identifying interventions to reduce wait times for primary care appointments supports our findings.¹² After abstracting data from 11 studies, they found that dedicated phone calls for follow-up consultations, along with other patient-centered interventions such as open access scheduling and e-consultations, were effective at reducing wait times.¹² Our project also showed that making scheduling changes in anticipation of possible no shows and late cancellations increased the available clinic slots for new patients. By considering the amount of past no shows and cancellations, two additional slots were added for new patients on the provider’s template.

Finally, our analysis suggests that follow-up appointments took up a considerable amount of clinic slots while discharges and as-needed appointments were minimally used. In this regard, clinic staff were educated to move toward discharging or assigning “as needed” follow-ups for existing patients with routine visits when deemed appropriate. This solution is in line with results from Potter et al’s study on breast cancer patients in the United Kingdom to address the burden of wait times, which concluded that reducing long-term follow-ups can be an effective method to increase clinic capacity and reduce wait time for new referrals.¹³ However, while results from their study initially showed a decrease in wait time for routine appointments, wait times subsequently rose as the number of referrals increased.¹³

Overall, the combination of reduction of no shows, early identification of cancellations, and elimination of unnecessary follow-up visits successfully decreased the clinic unutilized time and further freed clinic resources to accommodate more new patients, eventually decreasing our clinic’s waiting times for new patients.

This study has some limitations. First, the study only addressed simple interventions that were monitored over 3 months. Additional time for data collection may be warranted before results can be deemed sustainable in the long term and to identify the best processes to improve efficacy. Moreover, these processes were tested in only one provider’s clinic; other prospective studies and application to multiple centers will be needed to validate the results of these interventions on a larger scale. Lastly, while our pilot interventions proved effective in reducing wait times by a significant amount, they by no means addressed all sources contributing to the long wait times. Further optimization could be facilitated by tackling the problem of increasing service demand and limited staff capacity. For example, educating providers about the function and objective of pain clinics or embedding a mid-level provider triage via telephone or e-consultation could reduce inappropriate referrals; hiring more staff to increase pain clinic capacity could improve the issue of lack of resources.

In conclusion, limited resources, especially in community pain medicine specialty clinics, may cause long waiting times for new patients. Our project showed that simple modifications and changes in the clinic workflow can improve efficiency and time utilization, leading to a significant reduction in the waiting times for new patients. In the future, the results of this project and similar larger and more comprehensive projects may help to develop statistically validated analytics embedded in the electronic health records that can predict no shows and cancellations in real time.