A Czech national administrative real-world study of diagnostics and treatment pathways of non-small-cell lung cancer stratified by disease stage: From data to actionable indicators

This Czech national study utilized linked administrative and registry data to evaluate stage-stratified quality indicators for non-small-cell lung cancer care, revealing that despite improvements in multidisciplinary team discussions and centralization, fewer than half of patients initiated treatment within eight weeks and significant regional disparities in biomarker testing persist, leading to the implementation of these metrics as a national tool for continuous quality evaluation.

The Gist

Imagine the healthcare system as a massive, bustling train station. Every day, thousands of passengers (patients) arrive with a ticket to a specific destination: a cure for Non-Small Cell Lung Cancer (NSCLC).

This paper is essentially a traffic report for that station, written by a team of Czech data detectives. They didn't just count how many people got on trains; they tracked exactly how they got there, how long they waited, and whether they were sent to the right platform.

Here is the story of their journey, broken down into simple parts:

1. The Mission: Mapping the Map

In the past, the station managers only knew that people were sick. They didn't know exactly how sick they were (the "stage" of the cancer) or if the train they boarded was the right one.

The researchers combined two huge databases:

• The "Passenger Manifest" (National Cancer Registry): Tells them who has the disease and how advanced it is.
• The "Ticket Ledger" (Insurance Claims): Tells them what tests, surgeries, and medicines the patients actually received.

By gluing these two together, they created a high-definition GPS map of the entire journey for nearly 11,000 patients between 2017 and 2023.

2. The Four "Quality Checkpoints"

To see if the station was running well, they looked at four specific "Quality Indicators" (QIs). Think of these as the station's report card:

• The "Group Huddle" (MDT): Before a patient gets treatment, a team of experts (surgeons, oncologists, radiologists) should meet to decide the best plan.
  • The Finding: Only about 54% of patients got this group huddle. It's getting better over time, but it's like a popular coffee shop that still has a long line; not everyone gets to order their custom drink yet.
• The "Speed Limit" (Time to Treatment): Once the cancer is spotted, how fast does the treatment start? The goal is to start within 8 weeks.
  • The Finding: Less than half of the patients started treatment within 8 weeks. In fact, the wait times are actually getting longer, even though the system is trying to get faster. It's like a highway where traffic is getting worse even though they added more lanes.
• The "Specialist Hub" (COCs): Complex Oncology Centers are the "super-stations" with the best equipment and doctors.
  • The Finding: More people are being sent to these super-stations, which is good. However, it depends entirely on where you live. If you live in one city, you might get VIP treatment; if you live in a neighboring town, you might get stuck at a local stop.
• The "DNA Scan" (Biomarker Testing): For advanced cancer, doctors need to scan the tumor's DNA (like checking a lock to find the right key) to see if targeted drugs will work.
  • The Finding: This is improving. About 70% of patients with advanced cancer got this scan in 2023. But for early-stage cancer, it's still low (around 40%), mostly because surgery is often the main fix there.

3. The Plot Twists (Surprising Findings)

• The "Ghost" Patients: About 21% of the patients diagnosed with cancer never received any treatment at all.
  • Why? Most of these were very elderly or very sick, so the "train" was too fast for them. However, a small number were early-stage patients who could have been cured but were missed or refused treatment. These are the "missed connections" the station needs to fix.
• The "Missing Ticket" Problem: About 15% of patients were kicked out of the study because they didn't have a proper CT scan before their biopsy.
  • Analogy: Imagine trying to board a plane without a passport. The researchers said, "We can't trust the journey if we didn't see the ID check first." This suggests some patients are skipping crucial diagnostic steps.
• The "Slow Down" Paradox: The system is doing more things right (more group huddles, more specialist centers), yet the wait times are getting longer.
  • Why? The researchers suspect that because they are doing more complex tests (like the DNA scans mentioned above) and sending more people to distant "Super Stations," the logistics are getting complicated. It's like a restaurant adding a gourmet menu: the food is better, but the wait for the table is longer.

4. The Takeaway: Why This Matters

This paper isn't just a list of numbers; it's a blueprint for fixing the station.

The authors have built a tool that the Czech government can now use as a dashboard. Instead of guessing if cancer care is good, they can look at the dashboard and see:

• "Oh, the wait time in Region X is too long."
• "Region Y isn't doing enough DNA scans."
• "We need to get more patients to the Group Huddle."

In a nutshell:
The Czech Republic has built a powerful engine to track lung cancer care. They found that while they are getting better at sending people to the right specialists and doing advanced tests, the speed of care is lagging behind. The goal now is to use this data to clear the traffic jams, ensure everyone gets the "Group Huddle," and make sure no one misses their train to a cure.

Technical Summary

1. Problem Statement

Lung cancer remains the leading cause of cancer-related mortality globally, with Non-Small Cell Lung Cancer (NSCLC) accounting for approximately 85% of cases. In the Czech Republic, over 65% of new cases are diagnosed at advanced stages (III or IV). While quality indicators (QIs) for lung cancer care (e.g., timeliness, multidisciplinary team involvement) exist in various European countries, there is a significant gap in population-level data that integrates administrative claims with cancer registry staging information.

Previous studies often relied solely on administrative data (lacking staging) or registry data (lacking detailed treatment timelines). This study addresses the need for a comprehensive, stage-stratified analysis of patient pathways to identify delays, regional disparities, and adherence to standards of care, specifically focusing on the transition from diagnosis to treatment initiation.

2. Methodology

Study Design & Data Sources:

• Type: Retrospective, longitudinal cohort study covering the period 2017–2023, with follow-up data extending to September 2025.
• Data Integration: The study linked five distinct data sources using an encrypted national patient identifier:
  1. National Cancer Registry (NCR): Provided incident diagnoses, morphology codes (ICD-O-3), TNM classification, and clinical stage.
  2. National Registry of Reimbursed Health Services (NRRHS): Provided claims for procedures, dates, specialties, and ICD-10 diagnoses.
  3. Pharmaceutical Claims: ATC codes for drug administration.
  4. Hospitalization Records.
  5. National Death Registry.
• Cohort Selection:
  • Initial Pool: 15,886 incident NSCLC patients (ICD-10 C34) aged ≥18 with no prior malignancy.
  • Exclusions: Patients without prior CT/PET-CT before biopsy (to ensure diagnostic plausibility), those with treatment prior to the index date, or data discrepancies.
  • Final Verified Cohort: 10,669 patients (median age 69).
• Index Date Definition: The date of the first lung biopsy (BX) followed by histopathological examination (HP) within 60 days, confirmed by ICD-10 code C34.
• Key Quality Indicators (QIs) Assessed:
  1. Multidisciplinary Team (MDT) Discussion: Presence and timing of tumor board review.
  2. Time to Treatment: Interval from the last (PET)CT to the initiation of First-Line Therapy (FLT).
  3. Centralization of Care: Proportion of patients treated in Complex Oncology Centers (COCs).
  4. Predictive Biomarker Testing: Completeness of PD-L1 testing (analyzed specifically for 2021–2023 due to code availability).

Statistical Analysis:

• Descriptive statistics (medians, IQR, frequencies).
• Stratification by disease stage (I–IV), calendar year, and region.
• Statistical analysis performed using R version 4.4.

3. Key Contributions

• Methodological Innovation: Successfully integrated NCR staging data with NRRHS administrative claims to create a "verified" cohort, overcoming the limitations of using either source in isolation.
• Stage-Stratified Granularity: Unlike previous broad analyses, this study provides specific QI metrics broken down by disease stage (I, II, III, IV), revealing distinct pathway variations.
• National Implementation: The methodology has been adopted as a national tool for continuous quality evaluation of cancer care in the Czech Republic.
• New Indicator Proposal: Proposes PD-L1 testing completeness as a critical actionable QI alongside traditional metrics like MDT and time to treatment.

4. Key Results

Cohort Characteristics:

• Untreated Patients: 21.3% (3,380) of the initial cohort received no treatment. This group was predominantly advanced-stage (Stage IV: 65.5%) with a median survival of 1.6 months, suggesting poor performance status precluded active therapy.
• Diagnostic Gaps: 11.6% (1,837) were excluded due to the absence of CT/PET-CT prior to biopsy, highlighting potential gaps in initial diagnostic work-up quality.
• Stage Distribution (Verified Cohort): Stage I/II (27.6%), Stage III/IV (67.9%), Unknown (4.5%).

Quality Indicators Performance:

• MDT Utilization:
  • Overall: 53.9% of patients had an MDT review.
  • Trend: Increased significantly over time (e.g., Stage I–II rose from ~43% in 2017 to ~74% in 2023).
  • Median time to MDT: 37 days.
• Time to Treatment:
  • Critical Finding: Fewer than 50% of patients initiated treatment within the recommended 8-week window, regardless of stage.
  • Trend: The proportion of patients treated within 8 weeks declined over time, despite improvements in MDT and centralization.
  • Bottleneck: The primary delay occurred between imaging/CT and the MDT discussion, rather than between MDT and treatment initiation.
• Centralization (COCs):
  • Care centralization increased over time, particularly in advanced stages.
  • Regional Disparity: In 2023, the proportion of Stage IV patients treated in COCs varied widely by region (71% to 100%).
• Predictive Biomarker Testing (PD-L1):
  • Analyzed for 2021–2023.
  • Stage IV: 70.0% testing rate in 2023.
  • Stage III: 65.2% testing rate in 2023.
  • Early Stages: Lower rates (Stage I: 39.6%), consistent with surgical management dominance.
  • Significant regional variability in testing completeness was observed.

Treatment Patterns:

• Early Stages (I–II): Surgery (SX) was the dominant treatment (81% Stage I, 68% Stage II), though a slight decline in SX was observed, potentially replaced by Stereotactic Body Radiotherapy (SBRT).
• Advanced Stages (III–IV): Pharmacotherapy (PHT) was dominant. There was an increasing trend in neoadjuvant therapy (Stage II) and the use of immunotherapy/targeted therapy (Stage III/IV).

5. Significance and Implications

• Policy Impact: The study demonstrates that administrative data can effectively monitor real-world care pathways. The methodology is now embedded in Czech national legislation for quality evaluation.
• Identification of Systemic Delays: The finding that time-to-treatment is worsening despite better MDT utilization and centralization suggests a "trade-off." Increased centralization and more complex diagnostic requirements (e.g., mandatory biomarker testing) may be introducing logistical bottlenecks.
• Equity Concerns: Substantial regional disparities in MDT access, COC treatment, and biomarker testing highlight the need for standardized pathways to ensure equitable access to precision medicine across the Czech Republic.
• Clinical Practice: The high rate of untreated patients in early stages (Stage I/II) warrants investigation into whether these are missed opportunities for curative intervention due to comorbidities or referral delays.
• Future Directions: The study advocates for the inclusion of PD-L1 testing completeness as a standard QI and calls for further research into optimizing the "diagnostic interval" to prevent the decline in timeliness observed in recent years.

In conclusion, this study provides a robust, data-driven framework for evaluating NSCLC care, revealing that while structural improvements (centralization, MDT) are occurring, the ultimate goal of rapid treatment initiation is being compromised by systemic complexities, necessitating targeted policy interventions.