Chavez, Adriana L. PhD; Singh, Prachee PA-C, MS; Aleman, Andrea A. RN, OCN; Anderson, Jaime E. PharmD, BCOP; Jernigan, Colleen PhD; Ravi, Vinod MD
The Sarcoma Medical Oncology Center at The University of Texas MD Anderson Cancer Center evaluated 1538 new patients from September 2010 to August 2011. The time between the start of the new patient appointment and the finalization of a treatment plan had a mean of 5 days and a maximum of 66 days. This waiting period places an emotional burden on patients. Patients have reported that waiting for answers and information during cancer diagnosis and treatment was the most difficult part of their experience, and the waiting periods heightened their uncertainty and anxiety.1 In addition, cancer outpatients have expressed concerns associated with waiting times at almost every stage between prediagnosis and treatment.2
A finalized treatment plan and the quality of information provided at diagnosis are of great importance for the patient's experience. In a survey of 295 patients with cancer, 89% considered information about treatment plans important.3 In addition, the provision of a firm diagnosis reduces the potential anxiety, morbidity, and reattendance of patients.4 Patients with cancer who perceived that the amount of information given to them at diagnosis was “everything” reported lower levels of anxiety and higher satisfaction with treatment discussions.5 We believed that working to reduce the time to the finalization of treatment plans would result in more patients receiving all the relevant information about their fully finalized treatment plan on the day of their first visit.
Given the rarity of sarcoma, (accounting for <1% of the total number of cancer cases in the United States),6 patients often cannot find specialized treatment locally. Approximately 50% of our patients are nonresidents, and they incur lodging costs and other burdens of travel proportional to the time that they wait for their treatment plan to be finalized. In a survey of patients with cancer, more than half reported that traveling for treatments affected their ability to carry out their job or schoolwork.7
Through our process analysis, we found that failing to finalize the treatment plan during the first new patient visit also results in longer clinic visits, additional process steps in the clinic workflow, and longer time spent doing follow-up work.
At our center, the new patient process begins with the patient collecting his or her past medical records prior to the first appointment. Upon his or her arrival for the first appointment, the patient checks in, and his or her records are reviewed by a mid-level provider (MLP). The patient is then assessed by a registered nurse. The nurse assessment consists of measuring the patient's vital signs, conducting a pain assessment, and a review of symptoms. The patient is then interviewed and physically examined by an MLP. Finally, the patient is interviewed and educated about the natural history of the disease and treatment options by a physician. If deemed necessary to develop a treatment plan, the clinic team orders new images and laboratory tests and the patient obtains any pertinent missing past medical records. Before new tests can be scheduled, the patient's health insurance must be verified, which requires at least 48 hours. Once sufficient clinical data are available, the physician finalizes the plan of care and the pharmacist prepares the chemotherapy orders if appropriate.
This process results in the collection of sufficient data to finalize the patients' treatment plan, but it requires an average of 5 days from the new patient visit to be completed, and much longer in some cases. The purpose of this study was to reduce the time to the finalization of treatment plans to decrease the patients' burden of waiting.
A multidisciplinary team consisting of a physician, physician assistant, clinic nurse, pharmacist, center administrative director, and quality engineer was assembled with the goal of decreasing the median time between new patients' first appointment and the finalization of their treatment plans by 50%. We selected the median as our target measure because of the high variability in the baseline data.
Once our goal was identified, the team performed the following process improvement steps. (1) mapped the new patient visit and treatment plan formulation processes including time estimates for each step, (2) created a cause-and-effect diagram to identify and classify all the causes of delay, (3) created a plan to address the causes associated with the greatest delays on the basis of the findings of the cause-and-effect analysis, (4) implemented the plan and gathered postimplementation data, (5) compared the postimplementation data with data before the implementation of the plan, and (6) calculated the differences in our target measure before and after the implementation of the plan.
Data on the time to the finalization of the treatment plan were extracted from our Electronic Health Record system by the MLP. Preimplementation data were collected retrospectively for a convenient sample of all new patients who attended an initial appointment and received a finalized treatment plan during the 10 months prior to implementation, from September 2011 through June 2012. Cases with delays to the finalization of the treatment plan that were caused by conditions external to our institution's processes were excluded from our analyses. Postimplementation data were collected prospectively for all patients who had a finalized treatment plan from July 2012 through August 2012.
From the cause-and-effect diagram, the team identified the following 2 key causes of delays to the finalization of treatment plans: insufficient data at the time of the initial appointment (eg, past treatment records, diagnostic imaging, laboratory results) and delays in obtaining new diagnostic images (associated with the 48-hour interval for insurance approval that is required before scheduling imaging appointments and with limited appointment availability when scheduling on short notice). The process maps revealed that several steps and delays are created as a result of not having sufficient data at the time of the initial appointment to finalize the treatment plan, including additional coordination steps among the clinic team and waiting for past medical records to be acquired.
The team designed the following process changes to address the key causes of delay:
An MLP reviews the new patient's records 5 to 12 days before the initial appointment to identify missing records and order required tests.
A business center representative initiates the process of obtaining insurance authorization for the tests on the basis of a requisition by the MLP.
A scheduler arranges for the required tests to be administered within 24 hours of the patient's initial appointment.
A clinic representative calls the patient to confirm the appointment, instructing him or her to bring the missing records to the appointment and reminding him or her that the records may be necessary to develop a treatment plan.
To implement the process changes, cooperation was obtained by the clinical administrative director from the business center representatives to initiate the authorization process before the first patient visit, and an MLP took responsibility for prereviewing patients' records, with a physician serving as a tentative backup.
The study measure was the median time to treatment plan, defined as the median interval (measured in hours) between a new patient's first clinic visit and the time at which the treatment plan was finalized. Data on the time of finalization of the plan of care were collected considering the following different endpoints: for patients to be treated at MD Anderson, the endpoint was the time of the chemotherapy order signature; for patients to be treated elsewhere, the endpoint was the time at which chemotherapy recommendations were faxed to the patient's local oncologist or the time of the clinic note signature or telephone note signature; and for patients who did not need chemotherapy or had insufficient data to make a treatment decision, the endpoint was the time of the clinic note signature or telephone note signature. Statistical analyses were performed using Stata version 11.2 (StataCorp, College Station, Texas), considering a 95% confidence level.
Baseline data were collected retrospectively over a 10-month period and included 75 new patients; 2 outliers qualified for exclusion. Postimplementation data were collected prospectively over a 2-month period and included 16 new patients. The time to the finalization of the treatment plan ranged from 0.67 to 1585 hours at baseline, and the postimplementation range was 1.8 to 166.9 hours. The mean time to the finalization of the treatment plan was 123.6 hours at baseline and 49.1 hours postimplementation. The median time to the finalization of the treatment plan was 70.5 hours at baseline and 7.6 hours postimplementation. An 89% reduction in the study measure was achieved using the median because of the high variability in the output data. Figure 1 displays the change in the median time to the finalization of the treatment plan, measured before and after the process improvement implementation.
The change in median time to the finalization of the treatment plan was found to be statistically significant (P < 0.05), as shown in the Table. We stratified by process endpoints because each reflects inherent differences in the processes that result in significantly different intervals. For the endpoints that had enough observations, (clinic note and telephone note), the confidence intervals are statistically significantly different.
Table TIME TO FINALI...Image Tools
Minimizing variation was not a stated objective; however, the process changes resulted in a reduction in the variation of the study measure. This is illustrated in Figure 2 by the narrowing of the control chart limits; these were set to be 3 standard deviations above and below the average using the statistical software Minitab version 16.2.4 (Minitab Inc, State College, Pennsylvania).
A secondary finding was that when the treatment plan was finalized during the initial visit, there were fewer process steps compared with cases in which there were insufficient data during the initial visit to finalize the treatment plan. Process steps eliminated in the former case included instructing the patient about how to obtain the necessary data and discussing the treatment plan with the patient in a subsequent visit or phone call. The reduction in process steps resulted in time savings of 88 minutes for both the MLP and the physician (as calculated from process map time estimates) for the cases in which the treatment plan was finalized on the day of the initial visit.
On the basis of our findings, we can conclude that our process changes designed to have all the data necessary to finalize the treatment plan available at the time of the initial appointment significantly reduce the time to the finalization of the treatment plan. In addition, the changes implemented consisted of work being shifted to earlier in the process and did not result in additional work for the clinic team, with the exception of 1 phone call made to the patient.
Through quality improvement methods, the team was able to identify that a lack of sufficient data at the time of the new patient visit was the main cause of delays. The need for detailed past medical data to finalize a treatment plan is caused in part by the complexity of the disease (the World Health Organization classification of tumors currently lists more than 50 different histopathologic subtypes of soft tissue sarcomas)8 and cannot be avoided; hence, the importance of designing a process that results in complete data earlier on.
The large reduction in the median time to the finalization of a new patient's treatment plan has benefits to the patient, clinic team, and the institution. Patients with cancer have reported being frustrated by waiting while the necessary diagnostic tests and procedures were scheduled and performed and while results were analyzed before any diagnosis was made.9 In this study, we successfully minimized this wait time by preordering diagnostic tests as part of our process changes.
The time reduction achieved will likely result in emotional benefits for our patients. Waiting for a finalized diagnosis has been found to be frustrating and poorly understood by patients with cancer because their main goal was to begin treatment as soon as possible.9 In addition, more complete information in the form of a finalized treatment plan is valuable to patients. Of needs of patients with cancer during the prediagnosis phase, informational needs were identified as the most important in one multicenter survey, and these needs were reported to be greatest after patients had received their diagnosis.10 Therefore, our reduction of the time that patients wait for complete information about their treatment plans addresses important patient needs.
The cost of food and lodging for a non–local patient waiting for his or her treatment plan to be finalized is approximately $1600 (assuming a 2-person, 6-night stay). Between September 2012 and August 2013, approximately 42.6% of the patients seen in our center were from out of state (data extracted from our electronic health record system). Eliminating this multiday stay by finalizing the treatment plan on the day of the initial visit results in reductions in patients' wait, anxiety, and time away from home, in addition to the financial benefits. Although the clinical implications of delays in diagnosis and treatment have not been established, longer intervals between diagnosis and treatment have resulted in greater uncertainty and accompanying anxiety.11 Patient concern about delays between prediagnosis and treatment may relate to concerns that any delay will reduce the chances of a positive treatment outcome.2
The process changes we implemented had the secondary effect of reducing the time spent coordinating care by the clinical team, including time savings of 88 minutes for both the MLP and the physician (as calculated from process map time estimates). These time savings create opportunities to accommodate additional new patients and consult appointments, which currently have a 4- to 6-week waiting period. Each new patient generates a median of approximately $46 000 in downstream charges for the institution (extracted from institutional billing data from September 2010 through August 2011). Consequently, adding 1 new patient per month would increase annual charges by $550 000 per physician; the Sarcoma Medical Oncology Center staff currently includes 7 medical oncologists.
To our knowledge, there are no published reports of efforts to apply process changes that have successfully reduced the time to the finalization of a treatment plan. A program focused on 5 cancer types used quality improvement initiatives including process mapping to reduce the wait time to first treatment by reducing unnecessary delays and by prescheduling patients' care at convenient times; however, initiatives yielded mixed results.12 Another program was aimed at reducing the time from the diagnosis of 3 cancer types to the start of treatment by adding patient navigators to the staff.13 The addition of patient navigators led to improvements in the time to diagnostic resolution; however, this type of approach requires the addition of substantial staff resources, in contrast with our process changes that required minimal staff time investment.
Our work is relevant because patients treated at a tertiary care center are often nonresidents, and expedited treatment planning minimizes their expenditures and anxiety. In addition, our process changes yielded work and time savings for the clinic team, which is relevant for any oncology practice.
Using process improvement tools such as process maps and cause-and-effect analyses allowed our team to effectively identify targeted areas for reducing delays to the finalization of treatment plans. Some challenges to implementing the proposed changes include the time that must be programmed for the MLP to review patients' records days rather than minutes before their appointments and the need for a staff member to perform the additional task of calling the patients. However, if implemented, the proposed process changes could yield time and financial benefits for the patient, clinic team, and institution.
1. Fitch MI, Gray RE, McGowan T, et al. Travelling for radiation cancer treatment: patient perspectives. Psychooncology. 2003;12(7):664–674.
2. Paul C, Carey M, Anderson A, et al. Cancer patients' concerns regarding access to cancer care: perceived impact of waiting times along the diagnosis and treatment journey. Eur J Cancer Care. 2012;21(3):321–329.
3. Soothill K, Morris S, Harman J, Francis B, Thomas C, McIllmurray M. The significant unmet needs of cancer patients: probing psychosocial concerns. Support Care Cancer. 2001;9(8):597–605.
4. Newby D, Fox K, Flint L, Boon N. A' same day' direct-access chest pain clinic: improved management and reduced hospitalization. QJM. 1998;91(5):333–337.
5. Schofield P, Butow P, Thompson J, Tattersall M, Beeney L, Dunn S. Psychological responses of patients receiving a diagnosis of cancer. Ann Oncol. 2003;14(1):48–56.
6. Siegel R, Naishadham D, Jemal A. Cancer statistics, 2012. CA Cancer J Clin. 2012;62(1):10–29.
7. Lightfoot NE, Steggles S, Gauthier-Frohlick D, et al. Psychological, physical, social and economic impact of travelling great distances for cancer treatment. Curr Oncol. 2005;12(4):1–7.
8. Fletcher CD, Unni KK, Mertens F. Pathology & genetics: tumours of soft tissue and bone. World Health Organ. 2002;4: 12–18.
9. Pieters HC, Heilemann MV, Grant M, Maly RC. Older women's reflections on accessing care across their breast cancer trajectory: navigating beyond the triple barriers. Oncol Nurs Forum. 2011;38(2):175–184.
10. Wiljer D, Walton T, Gilbert J, et al. Understanding the needs of lung cancer patients during the pre-diagnosis phase. J Cancer Educ. 2012;27(3):494–500.
11. Baade PD, Gardiner RA, Ferguson M, et al. Factors associated with diagnostic and treatment intervals for prostate cancer in Queensland, Australia: a large cohort study. Cancer Causes Control. 2012;23(4):625–634.
12. Robert G, McLeod H, Ham C. Modernising Cancer Services: An Evaluation of Phase I of the Cancer Services Collaborative. Vol 43. Birmingham, England: Health Services Management Centre, University of Birmingham; 2003:17–38.
13. Raich PC, Whitley EM, Thorland W, Valverde P, Fairclough D. Patient navigation improves cancer diagnostic resolution: an individually randomized clinical trial in an underserved population. Cancer Epidemiol Biomarkers Prev. 2012;21(10):1629–1638.
health care quality; patient satisfaction; quality improvement