Article Tools
Quality in Action
Article Tools
Using Quality Improvement Methods and Time-Driven Activity-Based Costing to Improve Value-Based Cancer Care Delivery at a Cancer Genetics Clinic
Abstract
To meet increasing demand for cancer genetic testing and improve value-based cancer care delivery, National Cancer Centre Singapore restructured the Cancer Genetics Service in 2014. Care delivery processes were redesigned. We sought to improve access by increasing the clinic capacity of the Cancer Genetics Service by 100% within 1 year without increasing direct personnel costs.
Process mapping and plan-do-study-act (PDSA) cycles were used in a quality improvement project for the Cancer Genetics Service clinic. The impact of interventions was evaluated by tracking the weekly number of patient consultations and access times for appointments between April 2014 and May 2015. The cost impact of implemented process changes was calculated using the time-driven activity-based costing method.
Our study completed two PDSA cycles. An important outcome was achieved after the first cycle: The inclusion of a genetic counselor increased clinic capacity by 350%. The number of patients seen per week increased from two in April 2014 (range, zero to four patients) to seven in November 2014 (range, four to 10 patients). Our second PDSA cycle showed that manual preappointment reminder calls reduced the variation in the nonattendance rate and contributed to a further increase in patients seen per week to 10 in May 2015 (range, seven to 13 patients). There was a concomitant decrease in costs of the patient care cycle by 18% after both PDSA cycles.
Value-based health care aims to achieve the best outcomes at the lowest costs.1 Through the Value in Cancer Care Task Force, ASCO has emphasized three elements of value in cancer care: clinical benefit, toxicity, and cost.2 Clinical cancer genetics services integrate the personal and family histories of patients to identify patients who are at risk for a cancer predisposition syndrome for genetic testing.3 With the growing array of genetic tests, these services will also be crucial in helping patients select evidence-based options to avoid costly treatments that do not improve outcomes.4
Given the value of genetic testing, access to it is an important surrogate quality measure of oncology practices.5 However, providing such access can be challenging due to shortage of trained personnel, cost of genetic tests, and lack of awareness among the public and medical community. To address this, the National Cancer Centre Singapore restructured in 2014 to form a Cancer Genetics Service (CGS). Headed by a clinical cancer geneticist, the service offers genetics education, complete family history taking, psychosocial assessment, elaboration of mutation risk, informed consent, as well as pretest and posttest counseling.3,6-9 As awareness of the service grew, CGS resources faced challenges in meeting increasing demand; only one clinical cancer geneticist was available for two 3-hour clinic sessions per week. A quality improvement (QI) project was thus initiated, with the aim to increase service capacity of the CGS clinic by 100% within 1 year without increasing direct personnel costs. In this article, we demonstrate how we increased value-based cancer care delivery using QI methods to improve access to clinical cancer genetics services, with a reduction in cost measured using time-driven activity-based costing (TDABC).
Patients with high-risk features suggestive of hereditary cancer were referred to the CGS clinic from other subspecialty disciplines across the health care system and the country. After a few months of operation, a study team consisting of a clinical cancer geneticist and clinic patient service staff was formed in April 2014 to review care-delivery processes. Detailed process mapping was done to understand patient-related workflow in the clinic (Fig 1) and to devise interventions to improve time-efficiency of current processes. Because cost was one of our considerations, TDABC was used to measure main drivers of direct personnel costs in our processes.10 We chose TDABC because it complemented process mapping and it had proven useful for several leading health care institutions.11,12 Additional information is available in the Appendix (online only).
FIG 1. Baseline process map. Colored boxes represent resources providing care. Numbers circled at the bottom of each box represent the number of minutes needed to complete activity. At decision nodes, the probability of each patient passing through a specific pathway is indicated by percentage value. PSC, patient services clerk.
In our baseline process map, the study team identified the most time-consuming processes as taking a family history and genetics education upon initial consultation. These tasks were also performed by the physician, our highest-cost resource. As a solution, we proposed substitution of the physician with a qualified but less costly resource to eliminate waste and improve workflow.13 A genetic counselor was thus hired in August 2014 to perform these tasks. We evaluated this change following the plan-do-study-act (PDSA) methodology.14 The primary outcome measure was the number of patients seen per week at the CGS clinic. These data were collected via our health care system’s automated clinic appointment system. Secondary measures were direct personnel cost of care delivery calculated by the TDABC method and appointment access times (ie, time from referral to first clinic visit and time from sample acquisition to disclosure of results).
The first PDSA cycle lasted from August 2014 to October 2014. Initial adaptations to CGS clinic processes leveraged rapid cycle feedback gleaned through the PDSA methodology. We found that workflow was most efficient with the genetic counselor as the first point of contact, who documented detailed relevant personal and family history for the subsequent physician consultation. As processes were further refined, our counselor assumed the majority of the genetics education tasks, as well as informed consent for genetic testing performed in follow-up visits.
For our second PDSA cycle, from November 2014 to January 2015, we decided to address the erratic and considerable nonattendance rates undermining the capacity of the CGS clinic. Preappointment telephone reminder calls were proposed to reduce variation and improve predictability of clinic capacity usage.13 It was hypothesized that this would also improve the primary outcome measure of the number of patients seen per week. This change was implemented in November 2014 and carried out manually by our patient service clerks. Data on nonattendance rates were also collected via the computerized clinic appointment system as an additional secondary process measure.
After the first two PDSA cycles, the study team sought to consolidate its findings. Detailed process mapping of clinic processes after the second PDSA cycle was done by the team together with the genetic counselor (Fig 2). The accuracy of this process map was corroborated against actual clinic waiting times and consultation duration. These data were captured manually by our patient service clerks over a period of 6 weeks for every patient seen at the CGS clinic between February 16, 2015 and March 30, 2015. Direct personnel costs of the processes after the second PDSA cycle were then calculated and compared with the baseline costs of care delivery.
FIG 2. Process map after second Plan-Do-Study-Act (PDSA) cycle. Colored boxes represent resources providing care. Boxes that represent process changes are marked by an asterisk. Numbers circled at the bottom of each box represents the number of minutes needed to complete activity described in the box. At decision nodes, the probability of each patient passing through a specific pathway is indicated by a percentage value. PSC, patient services clerk; Q&A, question and answer; VUS, variant of uncertain significance.
Between April 2014 and May 2015, a total of 251 patients were seen in the CGS clinic. After the introduction of a genetics counselor during the first PDSA cycle, the weekly number of patients seen at the clinic rose from two in April 2014 (range, zero to four patients) to seven in November 2014 (range, four to 10 patients) (Fig 3). This represents an increase in clinic capacity of 350%, surpassing our original goal of 100%.
FIG 3. Control chart with three sigma control limits of the weekly number of patients seen at the Cancer Genetics Service clinic at National Cancer Centre Singapore. Gold points represent significant shifts on the control chart. Control chart rules for stability analysis were the Montgomery rules.15 CL, control limit; UCL, upper control limit.
In the second PDSA cycle, implementation of preappointment telephone reminder calls reduced variation in the CGS clinic’s nonattendance rate (Appendix Fig A1, online only). This contributed to a further increase in the number of patients seen per week, to 10 in May 2015 (range, seven to 13 patients) (Appendix Table A1, online only). A lag in the rise of patient numbers was noted from November 2014 to December 2014. This could have been due to the year-end holiday season in Singapore. The rise in weekly patient numbers was not accompanied by an increase in the weekly average time from referral to first clinic appointment, which remained at a median of 26 days (range, 7 to 49 days) (Appendix Fig A2, online only). For patients who subsequently underwent genetic testing during the study period, the quarterly average time from sample acquisition to disclosure of results also decreased from 91 days (range, 56 to 126 days) in April 2014 to 42 days (range, 27 to 56 days) in April 2015 (Appendix Table A2, online only).
Data collected on clinic waiting times and consultation durations showed that estimated average times after the second PDSA cycle process map closely approximated actual variation (Appendix Table A3, online only).
Cost calculations using TDABC showed that direct personnel costs of the patient care cycle were lowered by 18%. Further breakdown showed that costs were lowered across both initial and follow-up clinic visits (Appendix Table A4, online only).
Value in health care is a balance between the outcomes of care and the cost to achieve those outcomes. Health care systems increase value when patient outcomes improve without increasing costs and when costs are reduced without worsening patient outcomes. Therefore, measurement of outcomes and cost is an integral part of delivering high-value health care. Access to care is an important outcome measure. Increased access alone does not translate into better health outcomes without value-based, high-quality interventions.16 However, cancer genetics testing is a value-based intervention. In this setting, increased access can also encourage positive-sum competition among providers to deliver better value.17 Although access to care is measured by many health care systems, most of the same systems are unaware of the true costs of delivering specific processes of care. This is because most systems choose to measure costs based on reimbursements rather than consumption of health care resources. TDABC represents a useful approach to focus measurement on resource consumption over the entire care cycle on the basis of the specific medical condition of a patient.
In this study, we show how QI tools, such as process mapping and the PDSA methodology, can be combined with TDABC to measure and communicate the value of changes in outpatient clinic processes. We demonstrated that we increased value by achieving the desired outcomes of more access and less cost. Substituting a genetic counselor for a physician to do the same task also shows that when we enable health care providers to work to the top of their licensure and training, we can free higher-cost personnel to provide other services and reduce process costs.
Although hiring additional personnel intuitively translates into higher upfront salary costs, appropriate utilization in the care-delivery pathway benefits the service by being able to handle far higher patient capacity in a cost-effective manner. A number of randomized controlled trials in well-described syndromes (eg, familial breast cancer) have found that genetic counselors are equivalent to physicians in leading counseling sessions across outcomes such as patient anxiety, satisfaction, and knowledge.18 Because genetic diagnostics will increasingly play a larger role in oncology as well as medicine in general, we see the potential for many more genetic counselors to assist with patient education.
Another finding was that manual preappointment telephone reminder calls improved efficiency. This intervention significantly decreased variation in the clinic nonattendance rate and thus reduced the risk of unused capacity, indirectly improving access. This is consistent with previous studies which, interestingly, also suggested that manual telephone calls are more effective than automated telephone calls in reducing nonattendance rates.19 In terms of economic costs, our data suggest a direct personnel cost of $1.35 for a 5-minute telephone reminder call (Appendix Table A5), which could potentially save $17.81 to $54.20 in wasted personnel resources for a missed appointment (Appendix Table A4). It is a simple intervention that can easily be replicated in all outpatient clinic services and would be especially cost-effective for those with high nonattendance rates.
Similar to other new advanced technologies, genetic testing can improve cancer care outcomes but also increase costs of care delivery.20 Coverage for genetic testing by insurance plans is poor, and most patients pay out of pocket.21 A significant number of patients seek financial assistance for genetic testing. These cost constraints can lead to inefficiencies in workflow, such as the need for a second clinic visit after needs-based financial counseling to determine subsidy level by our medical social workers. This opens value-added opportunities for TDABC beyond QI initiatives: namely, to develop novel reimbursement models that move toward value-based payments. With appropriate and regular data collection, TDABC may also help to understand cost drivers of day-to-day operations so that institutions can optimize resource utilization and reduce duplication and service fragmentation.
To achieve buy-in from leadership for similar studies, the authors recommend that teams contemplate how their activities fit into their institution’s overall strategy for delivering superior value to patients. Proposed interventions should aim to improve operational effectiveness or lead to a synergistic fit with key organizational activities for maximal positive impact.
Several limitations to this study should be noted. The TDABC method has inherent weaknesses, such as the challenges of measuring time, evaluation of underactivity, the assumption of homogeneity of activities, and the use of actual or standard costs.22 Our sample sizes were small and thus did not allow for meaningful statistical comparison of pre- and postintervention data. Finally, qualitative data of patient and referring provider satisfaction, as well as patient knowledge, have not been collected. The study team is looking at such data collection for future PDSA cycles.
However, given the significant improvements from our straightforward methods and interventions, other teams are encouraged to use similar approaches to help their clinicians intelligently reengineer clinical processes to deliver greater value to patients in need of clinical cancer genetics services and cancer care in general. This need is especially pressing as genetic information becomes increasingly important in cancer management and as the oncology community seeks to extend cancer care delivery to resource-limited settings.
Acknowledgment
Supported by the National Medical Research Council Transition Award (to J.N.) and the Community Cancer Fund by National Cancer Centre Singapore and Lee Foundation (to the Cancer Genetics Service). R.Y.C.T. and M.M.-D. contributed equally to this work.
Conception and design: Marie Met-Domestici, Alexis B. Guzman, Soon Thye Lim, Khee Chee Soo, Thomas W. Feeley, Joanne Ngeow
Administrative support: Soon Thye Lim, Khee Chee Soo, Joanne Ngeow
Provision of study materials or patients: Joanne Ngeow
Collection and assembly of data: Marie Met-Domestici, Ryan Y.C. Tan, Joanne Ngeow
Data analysis and interpretation: Ryan Y.C. Tan, Marie Met-Domestici, Ke Zhou, Alexis B. Guzman, Thomas W. Feeley, Joanne Ngeow
Manuscript writing: All authors
Final approval of manuscript: All authors
The following represents disclosure information provided by authors of this manuscript. All relationships are considered compensated. Relationships are self-held unless noted. I = Immediate Family Member, Inst = My Institution. Relationships may not relate to the subject matter of this manuscript. For more information about ASCO's conflict of interest policy, please refer to www.asco.org/rwc or jop.ascopubs.org/site/misc/ifc.xhtml.
No relationship to disclose
No relationship to disclose
No relationship to disclose
No relationship to disclose
No relationship to disclose
No relationship to disclose
No relationship to disclose
No relationship to disclose
| 1. | ME Porter, T Lee: The strategy that will fix health care Harv Bus Rev 91:50–70,2013 Medline, Google Scholar |
| 2. | LE Schnipper, NE Davidson, DS Wollins, etalAmerican Society of Clinical Oncology: American Society of Clinical Oncology statement: A conceptual framework to assess the value of cancer treatment options J Clin Oncol 33:2563–2577,2015 Link, Google Scholar |
| 3. | JN Weitzel, KR Blazer, DJ MacDonald, etal: Genetics, genomics, and cancer risk assessment: State of the art and future directions in the era of personalized medicine CA Cancer J Clin 61:327–359,2011 Medline, Google Scholar |
| 4. | YC Shih, PA Ganz, D Aberle, etal: Delivering high-quality and affordable care throughout the cancer care continuum J Clin Oncol 31:4151–4157,2013 Link, Google Scholar |
| 5. | ME Wood, P Kadlubek, TH Pham, etal: Quality of cancer family history and referral for genetic counseling and testing among oncology practices: A pilot test of quality measures as part of the American Society of Clinical Oncology Quality Oncology Practice Initiative J Clin Oncol 32:824–829,2014 Link, Google Scholar |
| 6. | S Syngal, RE Brand, JM Church, etalAmerican College of Gastroenterology: ACG clinical guideline: Genetic testing and management of hereditary gastrointestinal cancer syndromes Am J Gastroenterol 110:223–262, quiz 263,2015 Crossref, Medline, Google Scholar |
| 7. | AM Hirschberg, G Chan-Smutko, WF Pirl: Psychiatric implications of cancer genetic testing Cancer 121:341–360,2015 Crossref, Medline, Google Scholar |
| 8. | C Macaron, BH Leach, CA Burke: Hereditary colorectal cancer syndromes and genetic testing J Surg Oncol 111:103–111,2015 Crossref, Medline, Google Scholar |
| 9. | TA Rich, AH Woodson, J Litton, etal: Hereditary breast cancer syndromes and genetic testing J Surg Oncol 111:66–80,2015 Crossref, Medline, Google Scholar |
| 10. | RS Kaplan, SR Anderson: Time-driven activity-based costing Harv Bus Rev 82:131–138, 150,2004 Medline, Google Scholar |
| 11. | RS Kaplan, M Witkowski, M Abbott, etal: Using time-driven activity-based costing to identify value improvement opportunities in healthcare J Healthc Manag 59:399–412,2014 Medline, Google Scholar |
| 12. | KE French, HW Albright, JC Frenzel, etal: Measuring the value of process improvement initiatives in a preoperative assessment center using time-driven activity-based costing Healthc (Amst) 1:136–142,2013 Crossref, Medline, Google Scholar |
| 13. | G Langley, K Nolan, L Provost, etal: The Improvement Guide: A Practical Approach to Enhancing Organizational Performance 359,2009 (ed 2) San Francisco, California Jossey-Bass Publishers Google Scholar |
| 14. | MJ Taylor, C McNicholas, C Nicolay, etal: Systematic review of the application of the plan-do-study-act method to improve quality in healthcare BMJ Qual Saf 23:290–298,2014 Crossref, Medline, Google Scholar |
| 15. | DC Montgomery: Introduction to Statistical Quality Control 172–175,2001 (ed4). New York, NY Wiley Google Scholar |
| 16. | JP Souza, AM Gülmezoglu, J Vogel, etal: Moving beyond essential interventions for reduction of maternal mortality (the WHO Multicountry Survey on Maternal and Newborn Health): A cross-sectional study Lancet 381:1747–1755,2013 Crossref, Medline, Google Scholar |
| 17. | ME Porter, EO Teisberg: Redefining competition in health care Harv Bus Rev 82:64–76, 136,2004 Medline, Google Scholar |
| 18. | BJ Wilson, N Torrance, J Mollison, etal: Improving the referral process for familial breast cancer genetic counselling: Findings of three randomised controlled trials of two interventions Health Technol Assess 9:iii–iv, 1-126,2005 Crossref, Google Scholar |
| 19. | PE Hasvold, R Wootton: Use of telephone and SMS reminders to improve attendance at hospital appointments: A systematic review J Telemed Telecare 17:358–364,2011 Crossref, Medline, Google Scholar |
| 20. | NG Thaker, SJ Frank, TW Feeley: Comparative costs of advanced proton and photon radiation therapies: Lessons from time-driven activity-based costing in head and neck cancer J Comp Eff Res 4:297–301,2015 Crossref, Medline, Google Scholar |
| 21. | ME Robson, AR Bradbury, B Arun, etal: American Society of Clinical Oncology policy statement update: genetic and genomic testing for cancer susceptibility J Clin Oncol 33:3660–3667,2015 Link, Google Scholar |
| 22. | M Gervais, Y Levant, C Ducrocq: Time-driven activity-based costing (TDABC): An initial appraisal through a longitudinal case study JAMAR 8:1–20,2010 Google Scholar |
Time-driven activity-based costing is a variation of activity-based costing, where process costs are analyzed based on time consumption of constituent activities. Durations of each process step were first documented using a timesheet managed by the clinic administrator. Cost calculations for each activity were then made based on the associated personnel resource involved in a process step and their adjusted average hourly rate. This rate was calculated by taking the total annual salary and benefit expense for a particular personnel group divided by the annual number of work hours in a year. The average hourly rate was divided by 60 to calculate a cost-per-minute rate. The activity cost was the time elapsed multiplied by the cost per minute. For a patient moving through the process, the direct cost was calculated as the sum of all the activity costs they encountered in the pathway. Where the pathway reaches a decision node, the percentage probability of each arm was used in cost calculations.
The number of work hours in a year was standardized at 2,340 for all personnel based on the prevailing 45-hour work week at the National Cancer Centre. Subsequent calculations were based on the following estimated annualized salaries, including benefits costs (Table A5).
|
| Week | No. of Patients Scheduled | No. of Patients Seen | Nonattendance Rate, % |
|---|---|---|---|
| April 1, 2014 | 4 | 2 | 50 |
| April 15, 2014 | 2 | 0 | 100 |
| April 22, 2014 | 4 | 4 | 0 |
| May 6, 2014 | 2 | 2 | 0 |
| June 3, 2014 | 5 | 4 | 20 |
| July 8, 2014 | 4 | 1 | 75 |
| July 15, 2014 | 4 | 2 | 50 |
| July 22, 2014 | 3 | 2 | 33 |
| July 29, 2014 | 5 | 2 | 60 |
| August 5, 2014 | 5 | 3 | 40 |
| August 12, 2014 | 5 | 2 | 60 |
| August 19, 2014 | 6 | 4 | 33 |
| August 26, 2014 | 7 | 5 | 28 |
| September 9, 2014 | 6 | 5 | 17 |
| September 16, 2014 | 4 | 3 | 25 |
| September 23, 2014 | 7 | 5 | 29 |
| September 29, 2014 | 5 | 5 | 0 |
| October 7, 2014 | 11 | 8 | 27 |
| October 14, 2014 | 6 | 5 | 16 |
| October 21, 2014 | 9 | 4 | 55 |
| October 28, 2014 | 3 | 3 | 0 |
| November 4, 2014 | 5 | 4 | 20 |
| November 11, 2014 | 12 | 10 | 16 |
| December 2, 2014 | 8 | 7 | 12.5 |
| December 9, 2014 | 7 | 6 | 14 |
| December 16, 2014 | 15 | 10 | 33 |
| December 23, 2014 | 3 | 1 | 33 |
| December 30, 2014 | 7 | 7 | 0 |
| January 5, 2015 | 12 | 10 | 17 |
| January 12, 2015 | 8 | 4 | 50 |
| January 20, 2015 | 6 | 5 | 17 |
| January 26, 2015 | 13 | 10 | 23 |
| February 10, 2015 | 9 | 7 | 22 |
| February 16, 2015 | 11 | 9 | 18 |
| February 23, 2015 | 12 | 7 | 41 |
| March 2, 2015 | 11 | 8 | 27 |
| March 9, 2015 | 9 | 8 | 11 |
| March 17, 2015 | 9 | 9 | 0 |
| March 23, 2015 | 10 | 7 | 30 |
| March 30, 2015 | 17 | 12 | 29 |
| April 27, 2015 | 12 | 9 | 25 |
| May 4, 2015 | 15 | 13 | 13 |
| May 18, 2015 | 11 | 7 | 36 |
| May 26, 2015 | 14 | 10 | 28 |
|
| Quarter | Average Length of Time, Days (range) | No. of Patients |
|---|---|---|
| April 2014 to June 2014 | 91 (56-126) | 2 |
| July 2014 to September 2014 | 64 (42-98) | 6 |
| October 2014 to December 2014 | 86 (34-119) | 7 |
| January 2015 to March 2015 | 45 (17-63) | 16 |
| April 2015 to May 2015 | 42 (27-56) | 6 |
NOTE. A total of 39 patients proceeded to genetic testing during the study period; two patients were excluded from analysis due to incomplete records. Although a reduction in average time from sample acquisition to disclosure of results is noted, the results are limited by the small sample size and confounding factors (eg, test turnaround time).
|
| Week | Average Waiting Time Before Consultation, Minutes (range) | Average Length of Time With GC, Minutes (range) | Average Length of Time Between GC and Consultant, Minutes (range) | Average Length of Time With Consultant, Minutes (range) |
|---|---|---|---|---|
| February 16, 2015 | 7.5 (5-12) | 46 (30-57) | 6.5 (0-21) | 21 (7-32) |
| February 23, 2015 | 20 (5-30) | 44 (NA) | 5 (NA) | 8 (NA) |
| March 2, 2015 | 12 (4-19) | 38 (34-41) | 5 (1-7) | 16 (15-18) |
| March 9, 2015 | 14 (3-25) | 34.5 (29-40) | 5 (5-5) | 29 (18-40) |
| March 16, 2015 | 8 (7-9) | 45 (32-58) | 9.5 (8-11) | 30 (30-30) |
| March 23, 2015 | 11 (4-20) | 40 (35-48) | 7 (5-15) | 32 (30-35) |
| March 30, 2015 | 16 (7-32) | 45 (NA) | 2.5 (0-5) | 25 (NA) |
| Average | 12.6 (3-30) | 41.8 (29-57) | 5.8 (0-21) | 23 (7-40) |
Abbreviations: GC, genetic counselor; NA, not available.
|
| Time Point | Baseline, SGD$ | After Second PDSA Cycle, SGD$ | Percentage of Change (reduction after second PDSA cycle) |
|---|---|---|---|
| First clinic visit | 70.95 | 54.20 | 23.6 |
| Second clinic visit | 24.45 | 17.81 | 27.18 |
| Third clinic visit | 53.55 | 49.71 | 7.2 |
| Total | 148.95 | 121.72 | 18.28 |
Abbreviations: PDSA, plan-do-study-act; SGD$, Singapore dollar.
|
| Resource | Annual Salary, SGD$ | Average Hourly Rate, SGD$ | Cost Per Minute, SGD$ |
|---|---|---|---|
| Physician | 150,000 | 64.10 | 1.07 |
| Genetic counselor | 59,800 | 25.56 | 0.43 |
| Patient services clerk | 38,400 | 16.41 | 0.27 |
| Phlebotomist | 42,240 | 18.05 | 0.30 |
Abbreviation: SGD$, Singapore dollar.
FIG A1. Control chart with three sigma control limits of nonattendance rate shows a decrease in variation after institution of preappointment manual telephone call reminders in November 2014. Gold points represent significant shifts on the control chart. Control chart rules for stability analysis were the Montgomery rules.15 CL, control limit.
FIG A2. Control chart with three sigma control limits of weekly average time from referral to first clinic appointment (days). Control chart rules for stability analysis were the Montgomery rules. A total of 146 patients were referred to the cancer genetics clinic from April 2014 to May 2015. We excluded from analysis 15 patients due to missing referral date records and 17 patients due to postponement of appointments. The access time to the first appointment did not significantly increase despite a significant increase in the weekly number of patients seen. CL, control limit; UCL, upper control limit.
