Characteristics and In-Hospital Outcomes of Single-Vessel Coronary Disease Intervention: A Propensity-Matched Analysis of the National Inpatient Sample Database 2016–2020

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Dear Authors, the issue of this paper is important and the paper is well-presented. However, my major concern is lack of the clear rules to refer patients to single-vessel CABG vs PCI.

THE patients characteristics in Table 1 should contain data on the coronary artery that was operated on, LM, LAD, RCA, Cx. Without this data, there is a risk of bias. As in our center, most single-vessel CABG procedures are performed for off pump anastomosis LIMA- LAD in view of better long-term outcomes and graft patency. Rarely, single-vessel CABG is performed for the RCA, even less often for Cx and its side-branches due to limited access. Whereas, single-vessel PCI is done for a variety of coronary arteries. 

I believe that the Authors should provide information on the coronary artery type, and procedure: off pump, open, minimally invasive access.

Only then, it would be possible to weight and compare PCI VS CABG.

Author Response

Comment 1: Dear Authors, the issue of this paper is important and the paper is well-presented. However, my major concern is lack of the clear rules to refer patients to single-vessel CABG vs PCI.

THE patients characteristics in Table 1 should contain data on the coronary artery that was operated on, LM, LAD, RCA, Cx. Without this data, there is a risk of bias. As in our center, most single-vessel CABG procedures are performed for off pump anastomosis LIMA- LAD in view of better long-term outcomes and graft patency. Rarely, single-vessel CABG is performed for the RCA, even less often for Cx and its side-branches due to limited access. Whereas, single-vessel PCI is done for a variety of coronary arteries. 

I believe that the Authors should provide information on the coronary artery type, and procedure: off pump, open, minimally invasive access.

Only then, it would be possible to weight and compare PCI VS CABG.

Response 1: Thank you for pointing this out. We agree with this comment. However, due to the lack of data available for this study, we have not been able to obtain the suggested key information such as coronary artery type and procedure performed that would have helped this study to be more robust. We have thus added these limitations in the paper on page number 9, lines 305-345. However, we think this study still provides a small but valuable insight into this subject. We hope that the additional changes made to the limitation section helps to clarify this gap in data availability. 

Reviewer 2 Report

Comments and Suggestions for Authors

This study aims to analyze in-hospital complications and events following percutaneous coronary intervention (PCI) and coronary artery bypass grafting (CABG) for single-vessel coronary artery disease (CAD). However, it presents a series of methodological errors that lead to arbitrary conclusions. Its value is also limited by the fact that investigates a disease (single CAD) where the indications are clear (PCI over surgery )and only in terms of the short-term complications which are expected to be more after surgery than PCI. 

 

-The objective of the study is misleading (Lines 17-19)

“Few studies have analyzed in-hospital complications and events following each procedure [percutaneous coronary intervention (PCI) and coronary artery bypass grafting (CABG)] for single-vessel coronary artery disease (CAD). This study aims to fill this void using a large propensity-matched real-world database” 

 

However, the population was selected according to their “procedural codes” (Lines 65-67) and not their diagnosis.  

 

Consequently, the study does not compare patients with a single CAD. It compares patients who underwent single-vessel intervention which is a totally different concept. This population includes patients with multi-vessel CAD who were not able to undergo total revascularization or patients with LAD disease who for technical reasons (e.g. chronic occlusion) were referred for surgery. On the other hand, the PCI population includes all single-vessel interventions even if the patient had PCI in the past or is scheduled for a PCI in another vessel in the future. It may also include patients who underwent primary PCI during acute MI which is a population with no indication for surgery in the firsthand. This population may also have multi-vessel disease. 

 

-Furthermore, the authors include in the “single vessel disease” group the left main disease. This is very confusing firstly because “left main disease” is not considered “single vessel disease” and secondly because is treated as a single vessel PCI or multi-vessel CABG. This means that in this study there are patients with “left main disease” only in the PCI group.

 

-The whole discussion in the paper is also misleading because is focused mainly on studies analyzing the results of patients with single-vessel disease and left main disease and mainly on their long-term results. 

 

-All-cause death in the PCI group is 2,7% which is very high for single-vessel PCI. This needs an explanation (maybe a lot of these patients are emergency primary and PCIs and elective cases???). The same applies to the hospital stay of surgical patients which is extremely long (9.75 days) for a single vessel CABG.

 

-In the “clinical characteristics” important information is missing (like LV function and pulmonary hypertension). If this data is not available, it should be mentioned in the limitations section. 

 

Generally, the study intended to compare two different procedures in the treatment of single-vessel CAD but it ends up comparing two completely different populations based only on the treatment and not the initial diagnosis. In addition, single-vessel CAD is not a grey zone that needs further investigation. It is treated with PCI unless is technically not feasible, in which case is referred for surgery. It is also well known that surgery has more short-term complications and better long-term results compared with PCI. So, this limits even more the value of the study. The dilemma has always been whether the long-term benefits of surgery are enough to take the risk of short-term complications.  

Author Response

Comments 2: 

This study aims to analyze in-hospital complications and events following percutaneous coronary intervention (PCI) and coronary artery bypass grafting (CABG) for single-vessel coronary artery disease (CAD). However, it presents a series of methodological errors that lead to arbitrary conclusions. Its value is also limited by the fact that investigates a disease (single CAD) where the indications are clear (PCI over surgery )and only in terms of the short-term complications which are expected to be more after surgery than PCI. 

 

-The objective of the study is misleading (Lines 17-19)

“Few studies have analyzed in-hospital complications and events following each procedure [percutaneous coronary intervention (PCI) and coronary artery bypass grafting (CABG)] for single-vessel coronary artery disease (CAD). This study aims to fill this void using a large propensity-matched real-world database” 

 

However, the population was selected according to their “procedural codes” (Lines 65-67) and not their diagnosis.  

 

Consequently, the study does not compare patients with a single CAD. It compares patients who underwent single-vessel intervention which is a totally different concept. This population includes patients with multi-vessel CAD who were not able to undergo total revascularization or patients with LAD disease who for technical reasons (e.g. chronic occlusion) were referred for surgery. On the other hand, the PCI population includes all single-vessel interventions even if the patient had PCI in the past or is scheduled for a PCI in another vessel in the future. It may also include patients who underwent primary PCI during acute MI which is a population with no indication for surgery in the firsthand. This population may also have multi-vessel disease. 

 

-Furthermore, the authors include in the “single vessel disease” group the left main disease. This is very confusing firstly because “left main disease” is not considered “single vessel disease” and secondly because is treated as a single vessel PCI or multi-vessel CABG. This means that in this study there are patients with “left main disease” only in the PCI group.

 

-The whole discussion in the paper is also misleading because is focused mainly on studies analyzing the results of patients with single-vessel disease and left main disease and mainly on their long-term results. 

 

-All-cause death in the PCI group is 2,7% which is very high for single-vessel PCI. This needs an explanation (maybe a lot of these patients are emergency primary and PCIs and elective cases???). The same applies to the hospital stay of surgical patients which is extremely long (9.75 days) for a single vessel CABG.

 

-In the “clinical characteristics” important information is missing (like LV function and pulmonary hypertension). If this data is not available, it should be mentioned in the limitations section. 

 

Generally, the study intended to compare two different procedures in the treatment of single-vessel CAD but it ends up comparing two completely different populations based only on the treatment and not the initial diagnosis. In addition, single-vessel CAD is not a grey zone that needs further investigation. It is treated with PCI unless is technically not feasible, in which case is referred for surgery. It is also well known that surgery has more short-term complications and better long-term results compared with PCI. So, this limits even more the value of the study. The dilemma has always been whether the long-term benefits of surgery are enough to take the risk of short-term complications.  

Response 2: Thank you for pointing this out. We agree with this comment.

Firstly, we acknowledge that the phrasing of the study objective may have created some confusion regarding the selection of the study population. In response, we have revised the objective to accurately reflect that patients were selected based on procedural codes for PCI and CABG, rather than exclusively by diagnosis of single-vessel CAD. Page 1 lines 17-21 now read:
"Few studies have analyzed in-hospital complications and events following percutaneous coronary intervention (PCI) and coronary artery bypass grafting (CABG) procedures for patients who underwent these interventions for single-vessel coronary artery disease (CAD). This study aims to compare the outcomes of PCI and CABG in such patients using a large propensity-matched real-world database based on procedural codes."

As for the comments about CAD and single vessel interventions, we accept that this could introduce bias as it also contains populations with multi-vessel disease. As such, this limitation has been clarified in the limitations section (Page 9, lines 292-314), with the addition of the following lines:
"In addition, we acknowledge that his study compares patients who underwent single-vessel interventions, not necessarily those with isolated single-vessel coronary artery disease (CAD). The population may include patients with multi-vessel CAD who were either unable to undergo complete revascularization or had technical limitations, such as chronic occlusions, leading to surgery. Additionally, the PCI group could include patients with prior or planned PCI in other vessels or those who underwent primary PCI during acute myocardial infarction (MI), a group not initially indicated for surgery. Therefore, the study may include patients with more complex coronary disease profiles. "

In regards to the comment about LAD not being strictly a single-vessel disease, we have added a line within the limitations section which points this out. (Page 9, lines 313-316):
"This study also includes patients with left main disease in the single-vessel intervention group, although left main disease is generally considered more complex and not classified as single-vessel disease. This may introduce bias, as left main disease is treated differently in PCI (as single-vessel) versus CABG (as multi-vessel)."

In regards to the comment about the discussion section, we acknowledge this but stress that while every effort was made to compare our findings with similar studies looking at short-term outcomes, this wasn't always possible as there were not many studies comparing this aspect, hence the knowledge gap we hoped our study would fill. Furthermore, our study compared short-term outcomes because NIS only provides in-patient data, therefore longer-term follow-up data is not available.

In regards to the comment about high mortality rates and long hospital stays, we have added these sections to our paper:

Page 9, lines 336-342:

"In our study, we did find that the all-cause mortality rate of the PCI group was quite high (2.7%). This might be due to a large proportion of the PCI cases being performed as primary PCI for acute myocardial infarctions, or in emergency settings where the risk of mortality is naturally higher compared to elective procedures for stable CAD. Further subgroup analysis based on the urgency of the PCI procedure could help clarify the contribution of these emergency cases to the overall mortality rate in the PCI group. "

Page 8, lines 286-290

"Patients within our dataset were older, and had a high burden of comorbidities, which may have necessitated extended monitoring to manage surgical complications. Differences in post-operative care protocols between institutions or regions, as well as healthcare system differences in discharge practices, can also contribute to variations in hospital stay."

Lastly, the comment about missing clinical characteristics was also added to our limitations section. 

(Page 9, lines 342-344):

"It also does not include information regarding the left ventricular function or the presence of co-morbidities that could affect the results such as pulmonary hypertension."

We hope that the additional changes made to the limitation section helps to clarify this gap in data availability. However, we think this study still provides a small but valuable insight into this subject.

Reviewer 3 Report

Comments and Suggestions for Authors

This is a well performed study, for which the authors should be commended.

The authors rightfully acknowledge the missing data such as the artery involved (left main stem, LAD) since these are often the key in the decision making. The same remars applies to the type of stents or bypass (left internal mammary or saphenous vein). 

Can mid or long-term data be retrieved? 

It is interesting to know which proportion of patients needed urgent, emergent or salvage PCI/CABG, since this affects outcome. Does the "weekend" as preoperative / preprocedural parameter offers any clue? 

Author Response

Comments 3: 

This is a well performed study, for which the authors should be commended.

The authors rightfully acknowledge the missing data such as the artery involved (left main stem, LAD) since these are often the key in the decision making. The same remars applies to the type of stents or bypass (left internal mammary or saphenous vein). 

Can mid or long-term data be retrieved? 

It is interesting to know which proportion of patients needed urgent, emergent or salvage PCI/CABG, since this affects outcome. Does the "weekend" as preoperative / preprocedural parameter offers any clue? 

Response 3: 

Thank you for pointing this out. We agree with this comment. However, due to the lack of data available for this study, we have not been able to obtain the suggested key information such as coronary artery type and procedure performed that would have helped this study to be more robust. We have thus added these limitations in the paper on page number 9, lines 305-345. However, we think this study still provides a small but valuable insight into this subject. We hope that the additional changes made to the limitation section helps to clarify this gap in data availability. 

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

There is high risk of bias resulting in wrong data results interpretation 

Reviewer 2 Report

Comments and Suggestions for Authors

The paper is clearly improved and can be published in the present form