Real world evidence of acute interstitial lung disease-related hospital admissions infers complex, multifactorial association between social deprivation and 90-day all-cause mortality outcomes: data from the North West of England.

This multicentre retrospective study of nearly 1,000 acute interstitial lung disease admissions in the North West of England reveals that while the most socially deprived populations face a high burden of hospitalization, the direct association between geographical deprivation and 90-day all-cause mortality is complex and statistically inconsistent, suggesting a multifactorial relationship rather than a straightforward causal link.

The Gist

The Big Picture: A Medical Mystery in the North West

Imagine the North West of England as a giant, bustling neighborhood. In this neighborhood, there is a group of people suffering from a serious lung condition called Interstitial Lung Disease (ILD). Think of ILD like a "rusting" of the lung's air sacs; it makes breathing harder over time.

The researchers wanted to solve a specific mystery: Does where you live (specifically, how poor or deprived your neighborhood is) change your chances of surviving a hospital stay for this lung disease?

They looked at nearly 1,000 hospital admissions between 2017 and 2019. They didn't just look at the patients; they looked at the "zip codes" to see how deprived each area was, ranking them from 1 (most deprived) to 10 (least deprived).

The First Clue: Who is Rushing to the Hospital?

The Analogy: Imagine a fire station. You might expect that the people living in the most dangerous, fire-prone areas would call the fire station the most often.

The Finding: The study found that the "fire" (the lung disease flare-up) was indeed hitting the poorest neighborhoods the hardest.

• 32.7% of all the hospital admissions came from the poorest 20% of the population.
• What this means: People in the most deprived areas are getting sick enough to need the hospital much more often than those in wealthier areas. It suggests that when you are poor, your health struggles might be more severe, or you might not be getting help early enough, so you end up in the ER only when things are critical.

The Second Clue: The Survival Puzzle

Once the patients were inside the hospital, the researchers asked: "Does being from a poor area make you more likely to die within 90 days?"

This is where the story gets twisty. It's not a straight line.

The Analogy: Imagine a race where the runners are wearing different colored shirts based on their neighborhood.

• The Expectation: You'd think the runners in the "Poorest" shirts (Red) would finish last (die sooner) and the runners in the "Wealthiest" shirts (Gold) would finish first (live longer).
• The Reality: The results were messy.
  • The runners in the second-poorest group actually had the worst survival rates.
  • The runners in the poorest group (the Red shirts) actually did better than the second-poorest group.
  • The runners in the wealthiest group were somewhere in the middle.

Why is this confusing?
When the researchers tried to fix the numbers using advanced math (to account for missing data), the link between "being poor" and "dying sooner" basically disappeared. It became statistically insignificant.

The Takeaway: Once a patient is actually inside the hospital receiving care, their zip code doesn't seem to be the main reason they live or die. The "poverty penalty" seems to happen before they get to the hospital, not necessarily during their stay.

The Real Villains: What Actually Predicts Death?

If the neighborhood isn't the main factor once you are in the hospital, what is? The study found three clear "villains" that increase the risk of death, regardless of where you live:

1. Being Male: Men in this study were more likely to pass away than women.
2. Needing Oxygen Before Arrival: If you were already hooked up to an oxygen tank at home before you got sick, your lungs were already in very bad shape.
3. Low Lung "Efficiency" (TLCO): Think of your lungs like a sponge. If the sponge is dry and brittle (low TLCO score), it can't soak up oxygen well. Patients with this specific lung damage had a harder time surviving.

The "Missing Data" Problem

The researchers had a bit of a headache with their data. It was like trying to solve a jigsaw puzzle, but 40% of the pieces were missing.

• They didn't have lung function test results for many patients.
• To fix this, they used a computer program to "guess" (impute) the missing numbers based on patterns.
• The Result: When they used the "guessed" numbers, the link between poverty and death vanished completely. This suggests that the initial scary link between poverty and death might have been a mirage caused by incomplete information.

The Conclusion: A Call to Action

So, what does this all mean for the real world?

1. The "Doorstep" Problem: The biggest inequality happens at the front door. People in poor neighborhoods are getting sick and rushing to the hospital much more often. This suggests they need better support in their communities (better primary care, better education about the disease, more trust in doctors) so they don't end up in crisis mode.
2. The "Hospital" Reality: Once they are inside the hospital, the doctors treat them. The hospital care seems to level the playing field. Being poor doesn't automatically mean you will die faster if you are already in the system.
3. The Future: We need to stop just looking at the "zip code" as a measure of poverty. A rich person can live in a poor area, and a poor person can live in a rich area. Future studies need to look at the individual's actual life situation, not just their address.

In a nutshell:
This study tells us that poverty makes people get sick and rush to the hospital more often, but once they are there, the hospital care helps smooth out the differences. The real battle is to get people the help they need before they have to call an ambulance.

Technical Summary

1. Problem Statement

Social deprivation is a well-established determinant of health, known to worsen outcomes in chronic diseases, including Interstitial Lung Disease (ILD). Previous registry data (specifically the British Thoracic Society IPF registry) indicated that higher social deprivation correlates with worse long-term survival in Idiopathic Pulmonary Fibrosis (IPF). However, the impact of social deprivation on acute ILD-related hospital admissions and short-term (90-day) mortality outcomes remains unknown. Given that the North West of England has the highest rates of deprivation in the UK, there is a critical need to understand if geographical deprivation influences acute admission patterns and mortality in this specific high-risk population.

2. Methodology

• Study Design: Multicentre, retrospective, observational study.
• Setting & Population: Data was collected from 11 NHS secondary and tertiary hospitals across the North West of England. The study included adults (≥18 years) admitted between January 1, 2017, and December 31, 2019.
• Data Sources: Admissions were identified via ICD-10 coding (e.g., J84.1, J84.8, J84.9) and verified through manual review of medical records, coding, and multidisciplinary team (MDT) documentation.
• Exposure Variable (Social Deprivation): Geographical deprivation was determined using the 2019 English Indices of Deprivation (IMD). Postcodes were converted to deprivation deciles (1 = most deprived, 10 = least deprived) and grouped into five quintiles for analysis.
• Primary Outcome: 90-day all-cause mortality (time from admission to death).
• Statistical Analysis:
  • Descriptive: Comparisons between quintiles using ANOVA, Kruskal-Wallis, and Chi-squared tests.
  • Survival Analysis: Kaplan-Meier curves with log-rank tests.
  • Multivariate Modeling: Cox proportional hazards models were used to estimate Hazard Ratios (HR).
    • Complete Case Analysis: Included 11 variables (demographics, comorbidities, admission labs) but excluded variables with >15% missing data (FVC, TLCO, admission oxygen).
    • Pooled Analysis (Multiple Imputation): To address high missingness in key physiological variables (FVC, TLCO, oxygen), Multiple Imputation by Chained Equations (MICE) was used to generate 20 datasets. Results were pooled using Rubin's rules.
  • Interaction Analysis: An age × deprivation interaction term was tested to assess if age confounded the deprivation-mortality relationship.

3. Key Contributions

• First of its Kind: This is the first study to specifically examine the relationship between social deprivation and acute ILD-related admission outcomes, distinguishing it from previous long-term chronic disease studies.
• Real-World Evidence: Utilizes a large, multicentre dataset (999 admissions) from a region with high deprivation, providing robust real-world data on admission patterns.
• Methodological Rigor: The study employs advanced statistical handling of missing data (MICE) to include critical physiological markers (TLCO, FVC) often excluded in retrospective studies due to missingness, allowing for a more comprehensive multivariate analysis.
• Patient Involvement: The study design was validated by "Research Champions" (expert patients) from Action for Pulmonary Fibrosis, ensuring the relevance of the study design to the patient community.

4. Key Results

• Admission Burden: There is a high burden of acute admissions from the most deprived populations. 32.7% (327/999) of all admissions came from individuals in the most deprived 20% (Quintile 5).
• Demographic Disparities: The most deprived cohort (Quintile 5) was significantly younger and had a lower comorbidity burden (Charlson Comorbidity Index) compared to less deprived quintiles, despite having a higher admission volume.
• Unadjusted Mortality: Unadjusted survival analysis showed a non-linear relationship between deprivation and 90-day mortality. Surprisingly, Quintile 2 (low-intermediate deprivation) had significantly worse survival (52.3% mortality) compared to Quintile 5 (33.5% mortality, p=0.0001).
• Multivariate Modeling (Complete Case):
  • Deprivation showed a borderline significant association with increased mortality (HR 1.038, p=0.050), suggesting lower deprivation (higher decile) might slightly increase risk, though this is counter-intuitive.
  • Consistent Risk Factors: Male sex (HR 1.343) and pre-admission long-term oxygen therapy (HR 1.931) were strongly associated with increased mortality.
• Multivariate Modeling (Pooled/Imputed):
  • The borderline association between deprivation and mortality disappeared (HR 1.001, p=0.928) when physiological data (TLCO, FVC, oxygen) were imputed.
  • TLCO Impact: Lower Transfer Factor for Carbon Monoxide (TLCO) values became a significant predictor of increased mortality (HR 0.890, p<0.0001).
  • Consistent Risk Factors: Male sex (HR 1.518) and long-term oxygen (HR 2.770) remained the strongest predictors.
• Interaction: The interaction between age and deprivation was not statistically significant, indicating that the age difference between quintiles did not explain the mortality patterns.

5. Significance and Implications

• Complexity of Deprivation: The study concludes that while social deprivation heavily influences health-seeking behaviors (leading to a high volume of acute admissions from deprived areas), its direct impact on mortality outcomes once admitted is complex and multifactorial. The initial non-linear relationship suggests that factors other than simple geographical deprivation (e.g., disease severity, unmeasured individual factors) drive mortality.
• Clinical Practice: The findings highlight that the most deprived patients are accessing acute care more frequently but are not necessarily the group with the highest short-term mortality risk in this specific dataset. This challenges the assumption that deprivation linearly correlates with acute mortality in ILD.
• Policy and Future Research:
  • Community Outreach: There is a need for targeted community outreach to build trust and improve health literacy in deprived areas to prevent delayed presentation.
  • Data Granularity: Future research must move beyond geographical proxies (IMD) to capture individual-level deprivation markers (e.g., income, education, dental health) to better understand the true drivers of health inequality.
  • Service Design: New healthcare interventions must be designed to avoid widening inequalities, particularly as digital health expands, ensuring access for deprived communities.

Limitations: The study is retrospective with significant missing data (requiring imputation), potential coding errors, and a timeframe (2017–2019) that predates widespread antifibrotic use, potentially limiting generalizability to current treatment eras. Additionally, geographical deprivation may not accurately reflect individual socioeconomic status.