High Stakes in Healthcare: Uncovering Surgical Mortality Rates in New York Hospitals

When I first started working in healthcare, I was struck by how much variability there is in patient outcomes. Some hospitals consistently deliver excellent care, while others seem to struggle, especially with high-risk surgical procedures. This realization inspired me to explore an important question: which hospitals in New York are underperforming in terms of risk-adjusted mortality rates for high-risk surgeries?

Why THIS Project?

Having spent years in healthcare, my fascination with patient outcomes motivated me to dig deeper into the data surrounding surgical mortality rates. I’ve seen the impact surgeries can have on patients’ lives, particularly during my time at the UNM Comprehensive Cancer Research & Treatment Center. It was there that I understood how crucial it is to ensure patients receive the best possible care.

What Readers Will Gain

By reading this article, you’ll gain insights into the performance of hospitals in New York regarding high-risk surgical procedures. I’ll share key findings, the data I analyzed, and some surprising results that emerged from the project. You’ll also learn about the factors influencing these outcomes and what can be done to improve patient care.

Key Takeaways

• 79.5% of surgical procedures did not show significant differences in mortality rates compared to state averages.

• The expected rate of mortality is a strong predictor of the observed rate, explaining 65% of the variance.

• Esophageal resection is the procedure with the highest mortality rate across hospitals.

• The Roswell Park Cancer Institute is notably responsible for the elevated mortality rates for esophageal resections.

Dataset Details

The dataset I used comes directly from the State of New York and includes inpatient discharge data from hospitals between 2009 and 2022. It can be found here. With 25,116 rows and 17 columns, it contains vital information, such as hospital names, procedure names, and different mortality rates. Cleaning the data was essential, as I encountered many hospitals with the same ID but different names and some procedures that lacked sufficient data.

Analysis Process

To analyze the data, I conducted several steps: cleaning the dataset to ensure accuracy, visualizing key patterns, and using statistical methods to draw conclusions. One surprising aspect was how high the mortality rate for esophageal resections was compared to craniotomies. I had expected craniotomy to have the highest rate, but it was less than a quarter of the rate for esophageal resections. This discrepancy highlighted the stark reality of certain high-risk procedures.

Visuals and Insights

1. Statistical Comparison Table: The majority of procedures had risk-adjusted mortality rates that were not statistically different from the statewide average. However, esophageal resections skewed this data significantly.

   

   

2. Scatter Plot of Expected vs. Observed Rates: This visual showed that while 65% of the variance in observed mortality rates can be explained by expected rates, there were notable outliers. One hospital, NYU Langone Orthopedic, had surprisingly low observed mortality rates despite a high expected rate of pneumonia. Another hospital, Calvary Hospital Inc, had heart failure and pneumonia mortality rates that had an observed mortality rate over 800 but an expected mortality rate of less than 200. This could warrant further study.

   

   

   Outliers with high observed rate and one outlier with low observed rate but high expected rate. These were all from Calvary Hospital Inc.

3. Bar Chart of Procedures: This chart illustrated that esophageal resections had the highest average risk-adjusted mortality rate compared to other procedures (procedures were filtered to those having a greater mortality risk than was found statewide).

   

   

   Esophageal Resection Mortality Rate is by far the highest.

4. Bar Chart of Hospitals: This visual highlighted that only one hospital was responsible for the high mortality rates associated with esophageal resections, indicating a need for further investigation.

   

   

   Only one hospital came up when filtering for esophageal resection and greater mortality risk than statewide.

Main Takeaways

The results of this project underline the importance of understanding surgical risks and how they vary by hospital. It’s crucial for hospitals to be transparent about the risks associated with procedures like esophageal resections. Patients should be fully informed, and hospitals, especially Roswell Park Cancer Institute, should aim to align their outcomes more closely with statewide averages.

Conclusion and Personal Reflections

This project taught me a great deal about the complexities of healthcare data and the importance of rigorous analysis. I encountered several challenges, particularly in cleaning the dataset, but overcoming them deepened my understanding of data integrity. This experience has reinforced my commitment to improving patient outcomes through informed healthcare practices.

Call to Action

I’d love to connect with you on LinkedIn! Please leave a comment with your thoughts or questions. Let’s continue the conversation about improving healthcare outcomes together!