Response by Cram et al Regarding Article, “Utilization of advanced cardiovascular therapies in the United States and Canada: an analysis of New York and Ontario administrative data”

We appreciate comments by Al-Omran and colleagues and their interest in our paper;¹ they bring up several excellent points. Specifically, we agree that our reliance upon administrative data limits our ability to evaluate clinical indications, that the low prevalence of comorbid illness in our Ontario data likely reflects under-coding, and that the lower utilization of EVAR for women in Ontario relative to New York warrants further exploration.

The limitations of our study notwithstanding, our study was designed to advance international health system comparison with a focus on cardiovascular procedures. International comparative health systems research relies upon large representative data sets to compare utilization, treatment patterns, treatment intensity and patient outcomes between different countries. Historically international comparisons have relied upon aggregated data collected by organizations such as the World Health Organization or Commonwealth Fund;^2,3 but the reliance upon aggregated data results in such studies lacking the granularity needed to delve deeply into how different countries are treating well circumscribed diseases or conditions.⁴

We began our research nearly 6 years ago using population-based administrative data to compare care in the US and Canada to explore a range of conditions including hip fracture, spine surgery, and joint arthroplasty;⁵ these initial studies have provided important insights into how differences in priorities and healthcare organization ultimately translate into differences in access to and utilization of orthopaedic procedures.

In the current study we switch our attention to cardiovascular procedures, finding significantly higher utilization of EVAR, LVAD and TAVR in New York relative to Ontario.¹ While our finding of significantly higher utilization of all procedures in New York was expected and consistent with our prespecified hypotheses (see our study protocol on Open Science at: https://osf.io/5gbwz/), several other findings were surprising including the fact that higher utilization rates in New York relative to Ontario increased progressively with increasing patient age and higher in-hospital mortality in Ontario.

It is important to recognize that our results merely scratch the surface of the work that is required. Ultimately, as Al-Omran and colleagues allude to, it would be ideal to link population-based administrative data with population-based clinical registries to allow for a more nuanced approach to assessing clinical indications and patient outcomes for EVAR, LVAD and TAVR. Likewise it would be ideal to have data not only from the entirety the US and Canada, but also from other developed countries across the world. We were gratified to receive funding in 2019 from the US NIH (R01AG058878) in 2019 and established the International Health Systems Research Collaborative (https://projects.iq.harvard.edu/ihsrc/people) to allow us to expand our international comparative work to include not just the US and Canada but also England, Netherlands, and Israel. Our long term goal is to use international health system comparisons to provide policy makers, physicians, and patients with benchmarking data that can be used to policy and spending.

While health services research is now a thriving scientific discipline, we are only just beginning to understand the similarities, differences, and tradeoffs inherent in different nations’ health care systems. Empirical research is needed to guide policy makers in maximizing value and public health and also for benchmarking health system performance. Without such research, we are truly flying blind.