Authorship and contributorship Ulf Aasebø: main author, implemented the Lean project; Hans Henrik Strøm: participated in the writing process; Merete Postmyr: project manager, responsible for statistics, contributed in the writing process.
The Lean method as a clinical pathway facilitator in patients with lung cancer
Version of Record online: 15 NOV 2011
© 2011 Blackwell Publishing Ltd
The Clinical Respiratory Journal
Volume 6, Issue 3, pages 169–174, July 2012
How to Cite
Aasebø, U., Strøm, H. H. and Postmyr, M. (2012), The Lean method as a clinical pathway facilitator in patients with lung cancer. The Clinical Respiratory Journal, 6: 169–174. doi: 10.1111/j.1752-699X.2011.00271.x
Ethics This study has been performed in accordance with the ethical standards laid down in the Declaration of Helsinki.
Conflict of interest The authors have stated explicitly that there are no conflicts of interest in connection with this article.
- Issue online: 14 JUN 2012
- Version of Record online: 15 NOV 2011
- Accepted manuscript online: 26 SEP 2011 04:26AM EST
- Received: 21 November 2010; Revision requested: 11 July 2011; Accepted: 12 September 2011
- Lean method;
- lung cancer;
- patient pathway
Introduction: Patient flow during workup for lung cancer is often difficult to influence because of delay at many levels: patient delay, doctors' delay and waiting time for X-ray and CT scan, referral to specialist, and waiting time for chemotherapy and radiotherapy or surgery. A mean workup time of 4 months is not unusual.
Objectives: To improve quality and shorten the workup time for patients with lung cancer.
Methods: It was decided to employ a program designed by the Toyota car industry, the Lean process, as a tool to improve workup time. A Lean process implicates all levels of an institution with project and focus groups having workshops to analyse present status and suggest improvements.
Results: Since introducing the Lean project, we have decreased the workup time from a mean of 64 days to 16 days, and the median time from diagnosis to surgery was reduced from 26.5 days to 15 days.
Conclusion: It is feasible to improve patient flow for patients with lung cancer by employing the Lean method as a pathway instrument.
Please cite this paper as: Aasebø U, Strøm HH and Postmyr M. The Lean method as a clinical pathway facilitator in patients with lung cancer. Clin Respir J 2012; 6: 169–174.
Delayed time for diagnostic workup in lung cancer has proven difficult to influence. With an approximate 85%–90% 5-year mortality and a potential for cure, if diagnosed early (stages I and II), workup time may be of importance (1). A mean workup time of 4 months is not unusual (2) with both patient, general practitioner (GP) and specialist delay. Up to 20% of patients may change status to palliation during this time, although this is an issue of controversy (3–12). Patient flow in this setting is highly dependant on communication and coordination in the health-care system. Many programs have been created to improve patient flow. Along with other programs successfully adapted by the industry as quality improvement models, Lean thinking is one such instrument that has been increasingly implemented by health institutions in the clinical pathway process (13).
Lean is a philosophy initially introduced by the Toyota car industry in the 1950s and consists of five basic rules to enable the workforce teams to look at their environment and identify obstacles, alongside ‘value stream mapping’, to remove unnecessary steps in the process. These five rules, often referred to as the 5-S, are sort, set in order, shine, standardise and sustain, which in the health-care system translates into identifying activities that adds value for the patient, design a flowchart, eliminate unnecessary steps, make the process flow (facilitate and demand change), standardise the process and make continuous incremental improvements. There is also a tool for ad hoc evaluation in case of obstacles along the process. A Lean process demands multidisciplinary participation and involvement, and motivation by the whole institution from the ground floor and up (14).
In December 2008, our institution decided to employ the Lean model as a quality improvement instrument. Lung cancer was singled out as one of several clinical pathways. It was a goal to narrow the patient time from first radiological contact until treatment was initiated.
Patients and methods
The University Hospital of North Norway is a 400-bed teaching hospital with a lung department consisting of 19 beds with an oncological day unit for chemotherapy, attending to approximately 250 new cases of lung cancer a year.
The Lean organisation had an executive board in the hospital leadership, a project group including the leaders or key persons of the departments involved, and finally a focus group made up of the workforce ‘on the floor’ (medical staff at all levels, nurses, nurses aids, secretaries and radiotherapists, etc.). The participating departments were lung, oncology, radiology, thoracic surgery, pathology and GP. A project manager was overseeing the process. The project started in January 2009, and 40 single randomly selected patients' clinical pathways were retrospectively investigated, all diagnosed with lung cancer in the period 2006–2008. By the use of statistical process control, the data revealed no specific process control (SPC). This applied for time from first chest X-ray, first computed tomography (CT), bronchoscopy, treatment start with chemotherapy or radiotherapy, and time to surgery. The median time from chest X-ray to CT of the chest and upper abdomen was 10 days, median time from chest X-ray to referral and through diagnostic workup was 64 days, and wait time for surgery was 26.5 days.
The focus group had five workshops and dissected all levels of the organisation. The project group analysed and streamlined the suggestions for future implementation
The implementation of the Lean process is summarised in the algorithm (Fig. 1). This figure connects to a hyperlink so when activating a green box, all procedures at each step are visualised. When a GP puts ‘lung package’ on a request for X-ray, it means that lung cancer is not to be ruled out, and a chest X-ray is taken within 1–3 days (ideally within 24 h). If the suspicion is confirmed, the patient is referred to the chest physician on call by the radiologist the same day, and a CT scan of the thorax and upper abdomen is performed within 3 days. The chest physician becomes a patient-responsible doctor (attending doctor) for the remaining workup. A bronchoscopy is scheduled within 1 week, except in cases where the patient is on anticoagulant therapy when time must be allowed for phasing out the treatment. All blood tests are automatically performed without being specifically ordered, as they are standardised on the hyperlink when ‘lung package’ is requested. The goal was that the total workup time should not exceed 4 weeks (from the first chest X-ray is done). Time for start of therapy should not exceed 1 week thereafter. Ideally, most patients should have specific therapy within 2 weeks after the first chest X-ray was done. Causes for delay are typically when there is a need for positron emission tomography- (PET) CT scan, mediastinoscopy and endoscopic bronchial ultrasound or if a pulmonary nodule is to be followed according to guidelines (9, 15). Waiting time for surgery should not exceed 2 weeks. The project was started on 19 October 2009. The evaluation was set to be after 6 months, and the goal was that 85% of patients should have their workup within 28 days.
Statistical process control is a measurement of quality, a method for ongoing improvement of processes. It has a big potential as a tool to understand and drive changes in organisations. It has its foundation in the theory of variation (16, 17). The graphical method using ‘control charts’ (in this case, an I-chart) is the tool to see and understand the common and special causes of variation. Control limits provide additional tests to identify specific causes of variation to look more closely at whether data appear randomly distributed between the limits. Test #1 is indicated when a single point falls outside a control limit. Test #2 is indicated when eight or more points fall on the same side of the centreline. Test #3 is indicated by a trend of seven or more points in an ongoing line that either increases or decreases. The centreline is the mean of all the values (arithmetic mean). The chart contains an upper and a lower control limit; these coincides with ±3 standard deviations from the centreline. With a normal distribution of data, 99.73% of the points will be within the control lines.
The key quality characteristic and process variable (PV) are time-lapsed from first contact with specialised health care to a definite diagnosis, the diagnostic packet path, but also the time from definite diagnosis to start of adequate treatment. The PVs are defined as the time interval to chest X-ray, first CT scan, bronchoscopy, starting chemotherapy, radiotherapy or surgery (16, 17).
Not applicable as this is a quality improvement project.
The accrued patients were evaluated on 1 June 2010. In the follow-up period, 69 patients were investigated according to the ‘lung package’ deal. Eighteen patients were still under investigation on suspicion of lung cancer, whereas 33 patients had the diagnosis confirmed, and 18 patients had the diagnosis excluded. The workup time for the diagnostic packet path was reduced from a median of 64 days to 16 days (Fig. 2), and 82% had a cancer diagnosis confirmed within 28 days. Five of six delays from ideal time were explained by prolonged time to diagnosis (stopping anticoagulation, PET-CT examination necessary, etc.); one case was because of deviation from procedures. The median time for the diagnosis of lung cancer to be excluded was median 10 days (mean 33 days). This process is less predictable, shown by the difference between median and mean. By examining the deviations from schedule, 60% were protocol violations. The median time from chest X-ray to CT scan was reduced from 10 to 5.5 days. A CT scan was done within 24 h for 33% of the patients. The median time from diagnosis to surgery was reduced from 26.5 to 15 days (Fig. 3). Time to chemotherapy and radiotherapy was not significantly decreased, as the time span was short initially (6 days and 5.5 days, respectively). The use of charts shows how the process is flowing within the control limits over time. Small variation is an indicator of well-functioning flow in the patients' journey from one step to another in obtaining a diagnosis.
To improve patient flow is often difficult because it depends on several diagnostic and different departments to interact. The different specialties have their own routines and quality programs that need to be adhered to. A recent Cochrane review has concluded that clinical pathway programs are associated with reduced in-hospital complications and improved documentation without negatively impacting on length of stay and hospital costs (18). The review was based on 27 studies involving 11 398 patients. Very few studies (none included in the review), however, have looked at the clinical pathway of cancer patients, and most studies have apparently been motivated by an intention of reducing costs and length of stay, as well as a quality improvement tool. We have, for many years, tried to improve our workup time for lung cancer, particularly regarding an expedite referral to a chest physician after a pathological chest X-ray, indicative of lung cancer, without success until the introduction of Lean process. This study is, to our knowledge, the first published systematic attempt to reduce workup time for lung cancer patients in Scandinavia, even though similar projects have been orally presented at conferences. In an interventional study ‘The Time to Treat Program’, the investigators were able to reduce the median time from suspicion of lung cancer to referral to specialist consultation from 20 to 6 days in a cohort of 420 patients. The median time from specialist consultation to CT scans decreased from 52 to 3 days. The median time from suspected lung cancer to diagnosis was decreased from 128 to 20 days, results comparable with our study. Eventually, 33% were diagnosed with lung cancer (1). Yoria et al. has shown that in the US health-care system, the time between routine chest X-ray, diagnosis and treatment of lung cancer can vary widely and is not associated with disease stage or survival, with an image-treatment median interval of 76 days for public hospitals (19). Whether a delay in patient flow has an impact on patient survival remains to be proven. It is, however, important to have a standard follow-up regimen for pathological changes on chest X-ray and CT scans. Pulmonary nodules are particularly important, as they are prone to become ‘missed opportunities’ and as they may remain virtually unchanged over years. The solitary nodule often prolongs the patient flow and may be misinterpreted as negligent (20–22). The commitment to a patient pathway program has shown to be effective in shortening the workup time and particularly to expedite the time from a pathological chest X-ray to CT scan and diagnostic procedures while short circuiting the GP. As the lung doctor gets the patient ‘hot’ from the radiologist, diagnostic workup can be done more frequently in an outpatient setting rather than evaluating a patient via referral. It also secures the patient from being lost in case of changing primary care doctors or GPs on leave.
The implementation did not encounter any problems, and the initial scepticism from doctors evaporated.
The role of the focus group cannot be overestimated, as the sense of ownership of the Lean developed during workshops is instrumental in sustaining continued Lean improvement. In fact, the Lean effort from the department of radiology has stimulated the development of similar pathways to other tumour groups. It is really a matter of communication, and the Lean fast-track method is also suitable for implementation in smaller institutions; but, it has to be born in mind that the process is long-term and needs surveyance and revisions.
- 14Going ‘Lean’ streamlines processes, empowers staff and enhances care. Hosp Health Netw. 2006;80: 34–5..
- 16Improving Healthcare with Control Charts – Basic and Advanced SPC Methods and Case Studies. Milwaukee, WI, American Society for Quality (ASQ), 2003..
- 17Measuring Quality Improvement in Healthcare – A Guide to Statistical Process Control, Applications. Milwaukee, WI, ASQ, 2001., .