Making sense of conservable days

1. Introduction
2. What are conservable days?
3. Why are conservable days important?
4. What are the main reasons for conservable day challenges?
5. Where can data analysis help?
6. Conclusion

Introduction

Conservable days represent one of hospital management’s most challenging metrics: everyone watches them, few understand them, and many struggle to improve them despite significant effort. Poor performance could indicate operational inefficiency, complex patient populations, data quality issues, or methodological mismatch, and usually some combination of all four.

What are conservable days?

Every inpatient visit is “coded” with standardized diagnoses and procedure codes. From these codes, an Expected Length of Stay (ELOS) is calculated based on the patient’s condition, age, and case complexity using CIHI case-mix methodologies and historical utilization patterns. When Acute Length of Stay (ALOS) exceeds ELOS, the difference represents “conservable days”: excess days the patient theoretically should not have occupied an acute care bed.

Why are conservable days important?

Ultimately, conservable days are about better patient care. Ministries and health systems rely primarily on accountability mechanisms (funding implications and performance evaluations) to compel change. Whether this approach is optimal or not, it’s the reality hospitals navigate.

It’s a widely used perform metric: Conservable days serve as an imperfect but widely used measure of hospital efficiency. Leadership is routinely evaluated on this metric.

There are funding implications: Provincial funding models can partially fund hospitals based on theoretical bed days (ELOS) rather than actual days (ALOS), creating financial disincentives for excess length of stay.

When used properly, it leads to positive change: Conservable days can identify care delivery problems. Tracking them to their source may reveal inefficient care paths or systemic bottlenecks that, once addressed, improve patient outcomes (longer stays correlate with higher infection rates) and reduce unnecessary costs.

What are the main reasons for conservable day challenges?

1. Operational Inefficiencies

Operational factors can meaningfully contribute to conservable days, independent of clinical care quality. Common examples include:

• Discharge Planning Gaps: Lack of weekend discharge capabilities, delayed physician orders, or poor communication.
• Care Coordination: Inadequate coordination with community services, home care, or post-acute facilities.
• Workflow Inefficiencies: Delays in diagnostic imaging, consultant reviews, or medication reconciliation.

Identifying and addressing these systemic issues can accelerate appropriate discharges and reduce hospital burden while improving patient experience.

2. Limitations in Case Grouping Methodologies

Case mix groupers are inherently imperfect. While they normalize expected length of stay across large populations, they cannot fully account for clinical, social, and operational complexity of individual patients or regional differences in care delivery.

Cases deemed “too long” may be reclassified as atypical and excluded from calculations, even when excess days stem from inefficiencies rather than clinical necessity. Conversely, genuinely atypical cases may remain classified as typical.

These methodologies are more reliable in aggregate than at the individual case level, but hospital-wide conservable day results can still be greatly distorted by the accumulation of classification issues across many cases. This can be of particular concern for hospitals with complex or non-standard patient populations.

3. Data Quality Issues

As a data quality assurance company specializing in hospital coding, we know incomplete or inaccurate coded data directly affects conservable day calculations. While much attention focuses on coding accuracy for case weight (CMI) and funded volume eligibility, data quality also directly affects ELOS. If visit details that increase complexity (such as diagnoses and interventions) aren’t captured, calculated ELOS will be lower than appropriate, artificially increasing conservable days.

Poor coding quality stems primarily from incomplete physician documentation and resource constraints. Medical coders must extract information from large charts under significant productivity pressures, often without direct physician access for clarification. The most complex cases requiring complete coding are precisely those most difficult to code accurately, creating consistent under-capture of complexity.

4. Conservable Day Calculation Nuances

Multiple calculation methodologies exist, each with different implications:

• Netted vs. Unnetted: Determines whether hospitals receive “credit” for cases discharged before ELOS, offsetting cases with excess days.
• Case Exclusions: Inconsistent exclusion of atypical cases creates comparability issues.
• Ontario-Specific HIG Methodology: Ontario’s Ministry of Health uses HIG for ELOS calculations, which does not consider patient comorbidities. This means regions with higher proportions of complex cases may show artificially inflated conservable days compared to the national CMG+ methodology, which accounts for patient comorbidity.

Understanding formula and methodology are being used is essential for interpreting results accurately and avoiding misleading conclusions.

5. Clinical Care Path Variations

Physician practice patterns vary significantly. Some physicians may order more diagnostic tests, maintain patients longer for observation, or use different treatment protocols.

However, attributing conservable days to clinical practice variation requires extreme caution given all the confounding factors above. Apparent variations most often reflect legitimate clinical judgment about complex cases rather than inefficiency.

Concluding that physician care path problems exist requires rigorous, granular analysis that accounts for case complexity, comorbidities, and other confounding variables. This must be treated with the utmost sensitivity and positioned as collaborative quality improvement rather than physician performance evaluation.

Where can data analysis help?

Because conservable days are influenced by clinical complexity, operational constraints, data quality, and methodological assumptions, no single metric or dashboard can “solve” them. However, thoughtfully designed data analysis can reduce noise, isolate meaningful signal, and focus improvement efforts in ways difficult to achieve through anecdote or intuition alone.

Across all analyses, it is essential to examine trends over time rather than relying on single-period results. Longitudinal analysis helps distinguish structural improvement or deterioration from short-term fluctuation, while statistical techniques can filter routine variability and highlight changes more likely to be meaningful.

Below are several practical ways data and algorithms can help:

1. Identifying High-Impact Populations

Hospital-wide conservable day percentages often hide more than they reveal. A small number of patient groups frequently account for a disproportionate share of conservable days.

Rank patient populations by contribution to total conservable days. Patient populations can be defined by HIG/CMG+ or Program/Service/Unit. Look for groups with high conservable days per case along with sufficient volume to account for statistical noise. Deprioritize low-volume or statistically unstable groups that introduce noise.

Targeted analysis focuses improvement efforts where changes will meaningfully reduce bed pressure rather than spreading resources across the entire organization. Equally important is establishing baseline measurements before implementing any initiative. Without clear “before” data, hospitals cannot determine whether improvements actually worked or if changes reflect normal statistical variation.

Start with patient populations following standardized care paths (e.g., hip replacements, routine C-sections), where deviations from expected length of stay more clearly signal process issues. These procedures follow predictable recovery patterns, making bottlenecks easier to identify and address. Complex medical admissions (elderly patients with pneumonia, heart failure, and dementia, for example) have inherently variable courses where it’s much harder to separate legitimate clinical complexity from fixable process delays.

2. Detecting Structural Discharge Barriers

Not all excess length of stay is clinical. Patterns in the data often reveal systemic discharge friction, such as:

• Spikes in conservable days aligned with day-of-week or holiday patterns, suggesting discharge delays rather than changes in patient acuity.
• Recurrent accumulation of conservable days late in the inpatient stay without corresponding indicators of increased clinical intensity, consistent with delays in transitioning patients from acute to ALC designation.

Algorithms that examine where conservable days accumulate within the inpatient stay (early versus late) can help differentiate treatment-related delays from discharge bottlenecks, while correlation and other statistical analyses can further highlight the underlying operational factors contributing to those barriers.

3. Improving Coded Data Quality

Expected Length of Stay (ELOS) is calculated directly from coded clinical data. When diagnoses, interventions, or comorbidities are not fully captured, case complexity is understated, resulting in lower than appropriate ELOS values and artificially inflated conservable days. Improving coded data quality helps reduce these false signals, ensuring that conservable day metrics more accurately reflect true performance rather than documentation gaps.

This represents a data-driven opportunity that doesn’t require operational or clinical care changes. At 3terra, our DQA platform identifies these coding gaps systematically, helping hospitals distinguish between real operational issues requiring investment and data artifacts requiring better documentation.

4. Peer Comparison and Collaborative Learning

Analyzing performance in isolation makes it difficult to distinguish internal issues from broader system effects. Secure, anonymized comparison with truly comparable peer hospitals at the program or service level provides a more meaningful reference point than provincial averages alone.

When hospitals can see how similar organizations perform, data becomes a starting point for collaborative dialogue rather than judgment. Peer analysis helps identify where differences are real and encourages the sharing of operational and clinical practices that support better performance, enabling hospitals to learn from one another and collectively improve patient flow.

5. Understanding Where You Are Different

Given the inherent limitations of length-of-stay methodologies, some hospitals may appear to underperform not because of poor care delivery or data quality, but because standardized calculations do not fully account for local or population-specific factors. In these cases, the issue is not performance but fit between the methodology and the hospital’s patient mix.

Examining which drivers of ELOS are excluded or de-emphasized, such as comorbidities in HIG-based analyses, can help explain why conservable day performance may lag on paper. Several important confounding factors such as teaching hospital status or regional staffing limitations are not reflected in the ELOS calculations.

This type of comparative analysis provides critical context for hospital leaders, allowing them to separate true performance issues from methodological mismatch and to clearly, quantitatively explain to stakeholders when standardized models do not fully reflect reality.

Conclusion

Conservable days aren’t a single problem with a single solution. Your hospital’s performance may reflect discharge planning gaps, methodological limitations, incomplete coding that understates complexity, or all three. Attempting to improve without understanding this composition leads to misdirected effort such as operational initiatives that can’t overcome data problems or coding improvements that can’t address genuine bottlenecks.

Data analysis illuminates these complexities. Stratified analysis reveals which populations drive excess days. Coding quality assessment separates measurement from operational problems. Peer comparison provides realistic benchmarks. These approaches help allocate improvement resources where they will generate real impact.

At 3terra, we address conservable days from both angles: data quality and analytics. Our DQA platform first identifies coding gaps that artificially inflate conservable days, distinguishing measurement artifacts from genuine inefficiency. Beyond data quality, our analytical tools help track down true operational issues through correlation analysis that links conservable days to root cause factors, with drilldown capabilities to examine individual cases and understand exactly where excess days accumulate and why.

Contact 3terra to assess whether your conservable day challenges reflect data quality issues, operational inefficiencies, or both.