Purpose: As part of quality improvement, we aimed to improve the average time from diagnosis to treatment to 34 days or less to improve outcomes in patients with stage III non–small cell lung cancer.
Methods: A multidisciplinary team developed 2 Plan-Do-Study-Act (PDSA) cycles. A statistical process control chart was used to analyze outcomes. Interventions include streamlining referrals from pulmonology to oncology, early patient navigator involvement, and proactive imaging orders.
Results: Baseline number of days from diagnosis to treatment averaged 44 days, with a median of 41 days. After PDSA cycle 1, the average was 33 days, with a median of 30 days. After PDSA cycle 2, the average was 34.5 days, with a median of 29 days from diagnosis to start of treatment.
Conclusion: Protocol changes included early involvement of lung cancer navigators to streamline referrals from pulmonology to oncology. Collaboration between physician groups, navigators, and pulmonary office managers reduced the time between diagnosis and treatment initiation. Our standardization of a process directed by a navigation algorithm supported these improvements in time to treatment.
Lung cancer is the leading cause of cancer-related deaths among men and women in the United States, with non–small cell lung cancer (NSCLC) being the most common type, making up about 80% of lung cancer cases.1 A study conducted by the American Lung Association showed that 48% of lung cancer cases were not diagnosed until the tumor had already spread to other parts of the body, while only 23% of cases were diagnosed at an early stage when the tumor was still limited to the lungs.2 Some of the most important factors influencing lung cancer survivability are early diagnosis and early treatment, with 5-year survival marked at 63% when cancer is detected in the localized stage, 35% when diagnosed at the regional stage, and 8% when diagnosed at the distant metastatic stage.2 Additional delays in treatment may contribute to worsened outcomes by further extending this timeline. Although some studies have suggested that ≤45 days from diagnosis to treatment to be a reasonable goal, a definitive timeline has not been established.3
Per the RAND Corporation recommendations for patients with NSCLC, patients without a prior diagnosis of cancer with a solitary nodule (<3 cm) on chest x-ray or CT scan of the chest should be offered further diagnostics within 2 months of the initial study.4 Treatment should begin within 6 weeks of official diagnosis unless brain metastases have been identified, in which case treatment should begin within 2 weeks of diagnosis.4 A large-scale Medicaid claims study5 showed that diagnosis-to-treatment intervals <35 days were associated with improved survival for patients with localized disease and for those with distant disease surviving >1 year.5 This was not applicable to patients with distant disease surviving <1 year.5
Additional studies have shown that treatment timelines may have a greater influence on outcomes in patients with higher-stage disease.6 These studies suggest that patients with stage ≥II NSCLC should be prioritized for more timely treatment. However, times to diagnose and treat lung cancer often exceeds recommended intervals, with the median time to diagnosis ranging from 8 to 60 days and the median time to treatment ranging from 30 to 84 days.7 Finally, it should be noted that additional factors to consider in the timeliness of treatment may include the difference in treatment timelines between private and public hospitals as well as patient demographics and disease stage.8
From February 2023 through February 2024, the average and median time from diagnosis to treatment with stage III NSCLC at Tallahassee Memorial Healthcare Cancer Center (TMHCC) was 43 and 40 days, respectively. At baseline, 45% of patients exceeded RAND Corporation’s proposed criteria of 42 days, and 58% exceeded the average time of 34 days from diagnosis to treatment, potentially resulting in less than favorable outcomes. While early-stage detection is strongly associated with improved survival, the benefits of diagnosis are evident only if it is followed by timely initiation of therapy. In other words, lung cancer survival is influenced not only when cancer is detected but also by how efficiently patients transition from diagnosis to treatment. Delays during this interval may allow for disease progression or missed opportunities for curative therapy. Thus, in addition to promoting early detection, minimizing time from diagnosis to treatment represents a critical and modifiable target for improving outcomes. Our quality improvement (QI) study aimed to decrease the average from 43 days to 34 days from diagnosis to treatment in a population of patients diagnosed with stage III NSCLC at TMHCC. By identifying delays in the current treatment timeline and implementing a more streamlined process, we sought to directly address this vulnerable interval in cancer care delivery and optimize future patient outcomes.
Our study design used QI tools with weekly quality measure data collection from March 2024 through November 2024. Leadership representation of this project included the TMH Pulmonary Physician Partners, TMH Medical Oncology Physician Partners, and the Oncology Data Program. Investigators formally met in March 2024 and twice in May 2024 to discuss QI intervention strategies. This was a QI project with the intent to improve local processes within our institution, and therefore an institutional review board request was not initiated.
The setting was a 772-bed private, not-for-profit community hospital with a cancer center that served 21 counties throughout north Florida, south Georgia, and south Alabama and was located in Tallahassee, Florida. The interdisciplinary team for this project included pulmonologists, oncologists, a nurse manager, a nurse, and 3 third-year medical students. All patients with stage III NSCLC who underwent biopsies at TMHCC and planned to get treatment at TMHCC were included. Inclusion criteria were adult patients aged 18 years or older, diagnosed with stage III NSCLC at TMHCC, and who continued to seek care through the TMHCC. The main exclusion criteria were the presence of obtaining either a biopsy or treatment at another institution, preexisting lung cancer, hospice, and/or death.
An oncologist, a nurse manager, a registered nurse, and 3 third-year medical students conducted structured chart reviews via electronic health records (EHRs) for data collection to collect all appointment dates. The nurse manager of the oncology program was notified of all patients with a positive pathology report. Data, including the date of diagnosis (defined as date of biopsy), and the start date of treatment, were collected. Measures included the average and median days following a positive pathology report. For the purposes of this study, treatment data were collected regardless of treatment type (radiation, chemotherapy, or immunotherapy). Using this information, we constructed a run chart (Figure 1).
We utilized a flow chart to evaluate the current workflow from time of diagnosis to first treatment. This was complemented by a root cause analysis using a fishbone diagram to identify the underlying issues that could lead to delays in treatment initiation, including issues surrounding resources within TMHCC, procedures, scheduling, or the patients’ life circumstances. We measured the reasons for delays in treatment start days using count calculation methodology. Initially, we collected granular data as a part of this process measure from January 2022 through January 2023 using the cancer center registry (Figure 2). However, due to reporting delays, we opted instead to obtain a new baseline from February 2023 through February 2024 for the remainder of the study, as seen in the run chart in Figure 1.
Our data sources were physician surveys and staff interviews that were conducted as a one-time survey, the cancer center registry, as well as patient chart reviews. Survey outcomes were translated into a Pareto chart with associated survey questions. Using these findings, we constructed a priority matrix (Figure 3).
Using the Plan-Do-Study-Act (PSDA) framework, we incorporated 2 QI interventions based on our baseline findings: (1) early intervention by the lung cancer patient navigators for streamlined pulmonary-to-medical oncology referrals, and (2) provided comprehensive checklist for navigators to follow. This algorithm included imaging orders necessary for proper staging, therefore expediting time to treatment. We found that implementation of these strategies was simple and inexpensive. Other high-impact interventions, such as facilitating patient transportation, hiring additional physicians and nurses, and increasing infusion room or radiation oncology capacity, although impactful in our priority matrix, were not feasible with the resources we currently have.
First, the TMH Pulmonary Physicians group was informed of the average time of 14 days that patients with a positive lung biopsy had before receiving a referral to medical oncology. Each specialty had a designated patient navigator. Navigators were trained registered nurses who served as a single resource for patients’ needs, including navigating the healthcare system, understanding and following up with diagnosis and treatment appointments, identifying resources and support, and emotional coping, as well as bridging communication between patients and specialists. To reduce delays, navigator involvement was initiated immediately upon receipt of a positive pathology report. Navigators were responsible for ensuring timely follow-up after pathology results and expediting referrals to medical oncology. Trained oncology nurse practitioners began seeing established patients for their follow-up appointments to allow oncologists to see new patients in a timely manner. Second, the lung cancer patient navigator was provided with a patient checklist. The checklist included ancillary studies, such as head MRI and PET scans, chemotherapy education, CT simulation, and laboratory studies, to be completed prior to the initial consultation with the oncologist. The navigator coordinated with the TMH Pulmonary Physicians group’s office manager to ensure that patients were scheduled accordingly. Imaging was to be completed within 5 days of biopsy. Using these QI interventions, we created a new algorithm for the patient navigators to refer to (Figure 4) and provided them with a checklist of this algorithm for every patient.
We performed a quantitative data analysis of the average and median of days from diagnosis to treatment start date. This quantitative analysis was conducted for both baseline data and postintervention. We used granular data from January 2022 through January 2023, current baseline data from February 2023 through February 2024, and results after intervention were collected from March 2024 through April 2024 for the first PDSA cycle, and May 2024 through November 2024 for the second PDSA cycle.
We constructed a statistical process control (SPC) x̄ bar chart to determine our outcome measure, ensuring that our intervention was stable and effective (Figure 5). Baseline for the SPC chart began in November 2023 to demonstrate a yearlong pattern until the end of data collection in November 2024. Approximately 42% of patients arrived at their initial consultation without a head MRI or PET scan. There were 175 patients at baseline data from February 2023 through February 2024. The average number of days from diagnosis to treatment start was 43 days, with a median number of 40 days. The first PDSA cycle began in March 2024 and ended in April 2024 with a total of 17 patients. The average number of days from diagnosis to treatment start decreased to 33 days, with a median of 30 days.
The second PDSA cycle began in May 2024 and ended in November 2024 with a total of 40 patients. This cycle had an average of 34.5 days and a median of 29 days from diagnosis to treatment. We observed 3 extreme outliers within the second PDSA cycle. These were a special cause variation outside of our control and did not represent the new process. For example, one of the special cause variations was due to prolonged decision-making by the patient.
This preliminary study demonstrates that targeted QI interventions can lead to measurable improvements in reducing the time to start of treatment for patients with stage III NSCLC within a community hospital setting. Although much of the lung cancer literature emphasizes early detection as the primary driver of improved survival, our findings underscore an equally important facet of cancer care: the timeliness of treatment initiation after diagnosis. The survival advantage associated with early-stage detection may be undermined if substantial delays occur between biopsy confirmation and the start of therapy. Therefore, optimizing postdiagnostic workflows represents a critical opportunity to translate early diagnosis into significant clinical benefit. After implementing a coordinated diagnostic and referral process centered around early patient navigator involvement and proactive imaging orders, there was a reduction in both average and median time from diagnosis to treatment after both PDSA cycles. After PDSA cycle 2, we had a decrease in the average number of days to treatment from 43 to 34.5 days. We also found a decrease in the median number of days with a decrease from 40 to 29 days. These findings underscore the potential for low-cost, process-focused interventions to address systemic delays in cancer care delivery. Our root cause analysis identified that nearly half (42%) of patients arrived at their first oncology consultation without having completed head MRI or PET imaging, which is essential for accurate staging and treatment planning. This delay in imaging represented a critical barrier to timely initiation of therapy.
The QI intervention addressed this gap by equipping the lung patient navigator with a structured checklist and algorithm. The old workflow followed a linear sequence of actions where subsequent steps could not be continued until previous steps were completed. The new workflow, as demonstrated in Figure 4, allowed for navigators to approach patient-centered care in a proactive manner while simultaneously addressing other components of care (medical records, scheduling, referrals, etc). This fostered direct communication between oncology and the pulmonary team to ensure imaging orders were placed within 5 days of biopsy. The resulting workflow optimization enabled earlier staging and reduced time to treatment intervals.
Our results represent meaningful progress after 6 months of tracking data. These results suggest that even incremental improvements in workflow coordination can significantly impact care timelines particularly for patients with locally advanced disease where timely treatment is most critical. This is consistent with prior studies that have shown that shorter diagnosis to treatment intervals is associated with improved outcomes in early and locally advanced NSCLC.2,5
It is important to note that the observed improvement over baseline was mainly attributed to the earlier months following the intervention. After the first cycle, there was immediate improvement. The short period of time in PDSA cycle 1 may not have accounted for the variation that we saw in the subsequent cycle. For PDSA cycle 2, we had a learning period followed by common cause variation. These variations were due to transportation and insurance issues that led to delays in seeking and receiving treatment. We noticed a slight decline in the later months, with the last 2 months of data collection significant for more patients having taken greater than 34 days for treatment initiation. Our lung cancer patient navigator, who was trained in May 2024 upon intervention implementation, began extended leave starting in October 2024 with return toward the end of the year. During this time, a patient navigator from a different specialty service temporarily assumed the role and responsibilities. It is plausible that this disruption in continuity of workflow, in addition to less familiarity and training on these protocol changes, could have led to a decrease in efficiency of care coordination. Moreover, the decline may also have been attributed to smaller sample sizes during these last 2 months compared with other months, thereby limiting the impact of these protocol changes. Nonetheless, even in the presence of variation, the number of days between diagnosis and start of treatment was well below the goal in PDSA cycle 2.
It is noteworthy that throughout intervention implementation, there was only 1 lung cancer patient navigator involved. Balance measures were assessed for unintended consequences regarding increased workload for our navigator. The navigator voiced no concerns or complaints regarding job burden after these protocol changes, which included essential, but additional, responsibilities. With the navigator playing a core role in our intervention success, we anticipate sustainability in the absence of balancing measures from these added responsibilities. In addition, the intervention did not accrue any additional costs as we used the resources we already had, that is, our lung cancer patient navigator. The observed transient increase in the number of days to treatment highlights potential vulnerability in sustained success when relying on a single individual. Strategies to counter this may include cross-training patient navigators of different specialties or pulmonary administrative staff to fulfill the role in the event of an absence.
The transition from the old EHR system to the new EHR after the start of 2025 limited our ability to obtain data to complete balance measure longitudinal data. During this transition, we stopped tracking newly diagnosed patients, which inevitably led to data fragmentation from this time until the complete conversion to the new EMR in March 2025.
The success of our QI initiative underscores the feasibility, scalability, and accessibility of patient navigator– driven workflow optimization in an environment where resources are limited.
Moreover, this initiative was conducted at a community hospital setting with a relatively small cancer program and limited specialty resources. The success of our QI initiative underscores the feasibility, scalability, and accessibility of patient navigator–driven workflow optimization in an environment where resources are limited. Unlike large academic medical centers where dedicated care teams, innovation infrastructure, and specialized services are available, our intervention relied on pragmatic and cost-effective strategies. These strategies used existing personnel and infrastructure that have meaningful implications for other community cancer centers to improve timely access to care without incurring additional financial costs. Our modest change in expanding responsibilities to existing staff members highlights how low-cost interventions can lead to measurable improvements in a short time frame.
This QI initiative and results align with broader literature recommending a 35- to 45-day window from diagnosis to treatment as a reasonable benchmark for timely care.5,6 Our findings also reinforce that survival in lung cancer is not solely determined by stage at presentation but is also influenced by care coordination efficiency. Patients with stage III disease, who often require multimodal therapy and extensive coordination, may be especially vulnerable to delays. Delays can contribute to tumor progression or declining performance status, potentially rendering patients ineligible for curative treatment. The biggest components of success in this project were the algorithm and early navigation involvement, including decreasing time from diagnosis to referral and scheduling. Our findings further support the need to prioritize patients with stage ≥II disease in future workflow improvements.3
This study was conducted at a single institution, limiting generalizability. Patient navigators were monumental in progressing this intervention as well as streamlining referral and imaging orders. This may not be generalizable or equally efficient in other institutions without equivalent staff support. The sample size of the postintervention group in both cycles remains relatively small, and additional longitudinal data are needed to determine the sustainability of the observed improvements. Moreover, as a retrospective QI project, our analysis was limited to administrative and clinical timelines. We did not evaluate treatment outcomes, such as progression-free survival, which is critical for fully assessing clinical impact. Lastly, factors such as socioeconomic status, comorbidities, and patient-specific delays were not evaluated in detail and may have contributed to variability in treatment start times. Despite these limitations, our findings highlight the importance of early imaging and interdisciplinary coordination in reducing time to treatment for patients with stage III NSCLC.
Future directions include expanding the navigator-led workflow to other stages and cancer types, incorporating patient-reported outcomes, and further evaluating the downstream effects on survival and quality of care. Continued attention to workflow optimization and care navigation will be essential in improving timely access to cancer treatment across healthcare settings. We found that treatment for uninsured patients was delayed compared with their insured counterparts. During our second cycle, we experienced significant variation from identifiable causes, such as transportation and insurance issues, that will benefit from an improvement in future processes. Future directions include another PDSA cycle that identifies financial toxicities in patients who are uninsured and performs a financial distress questionnaire (FACIT COST) to determine barriers to their care. This may help facilitate the approval of treatment plans and address transportation and treatment barriers. Navigators may consider collaboration with other hospital services, such as social services or patient advocates, to enroll patients in Medicaid and/or refer them to WeCare (a nonprofit organization of volunteer physicians who provide specialty medical care to low-income, uninsured patients within counties of north Florida). Finally, providing transportation may assist in earlier treatment initiation for this demographic.
These interventions led to improvement in the average, median, and percentage of patients who began treatment ≤34 days after diagnosis of stage III NSCLC. Protocol changes included early involvement of lung cancer patient navigators to identify those in need of imaging prior to their first consult. Collaboration between navigators and pulmonary office managers led to a streamlined referral process that ensured prompt referral and scheduling. This expedited any ancillary studies prior to the initial consultation. We established a new process at our institution and will continue to improve this through additional PDSA cycles addressing financial and transportation barriers to care.
The authors report no conflict of interests.
This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.
MW: Writing (original draft, data curation, formal analysis), review and editing, and visualization. AS: Writing (original draft, data curation), review and editing. KB: Writing (original draft, data curation), review and editing. JV: Data curation. LR: Data curation. JPDA: Conceptualization, methodology, formal analysis, writing, review and editing, and supervision.
Poster presented at the Florida Medical Association conference on July 26, 2025.
Keep up to date with the latest news from us via social networks:
To sign up for our print publication or e-newsletter, please enter your contact information below.
