Review Reports - Early vs. Late Readmission following Percutaneous Coronary Intervention: Predictors and Impact on Long-Term Outcomes

Round 1

Reviewer 1 Report

1) The abstract should be revised, as it is not mentioned any comparison between early and late readmission following PCI (as was mentioned in the title).

2) The overall readmission rate at 1 year was relatively lower (7.2%) as compared to other mentioned studies in the manuscript (18.6% - 50.4%). Therefore, this issue should be discussed, including some possible explanations.

3) The authors sated: “The causes of unplanned readmission beyond 30 days after PCI and the long-term implications of such readmissions are poorly characterized”. However, there were published some studies on large cohorts of patients, investigating incidence and predictors for unplanned readmissions following PCI (https://pubmed.ncbi.nlm.nih.gov/33050476/,https://pubmed.ncbi.nlm.nih.gov/30928446/).

4) The methods sections should be more detailed, including eligibility criteria and definition of outcomes (e.g., MACE).

5) The authors investigated predictors for unplanned readmissions following PCI and subsequent outcomes. However, planned PCI (10.9%) was mentioned as a cardiac cause of unplanned readmissions. Therefore, the results should be revised.

6) Although the authors also aimed to investigate predictors of unplanned readmissions after PCI, conclusions (in the abstract and in the main text) did not reflect any variable linked to a higher risk (or lower) of readmissions. Thus, conclusions should be revised, to highlight the main findings according to manuscript objectives.

Author Response

Reviewer #1 Comments:

We thank the reviewer for their comments and thoughtful review.

1 ) The abstract should be revised, as it is not mentioned any comparison between early and late readmission following PCI (as was mentioned in the title.

Response:

We thank the reviewer for highlighting areas that will benefit from clarification and expansion

Action:

The abstract has been completely revised to reflect additional analyses detailed within the manuscript concerning the comparison between characteristics and outcomes of patients with Early and Late readmission within the first year post-PCI. The abstract now focuses on explaining the methods and analyses, which include a comparison between firstly the overall group of patients with unplanned readmission vs. no readmission in the first year post-PCI, and secondly between patients who experience Early vs. Late readmission within the first year post-PCI. These changes can be found on page 1.

2 ) The overall readmission rate at 1 year was relatively lower (7.2%) as compared to other mentioned studies in the manuscript (18.6% - 50.4%). Therefore, this issue should be discussed, including some possible.”

Response:

We thank the reviewer for highlighting this very important point

Action:

We have expanded the discussion in paragraph 2 to address potential explanations for the cohort’s lower 1-year readmission rate compared to some other series. These changes can be found on page 9.

3 ) The authors stated: “The causes of unplanned readmission beyond 30 days after PCI and the long-term implications of such readmissions are poorly characterized”. However, there were published some studies on large cohorts of patients, investigating incidence and predictors for unplanned readmissions following PCI (https://pubmed.ncbi.nlm.nih.gov/33050476/,https://pubmed.ncbi.nlm.nih.gov/30928446/).

Response:

We agree with the reviewer that the 2 reports referenced assist in understanding predictors of unplanned readmission after PCI, however, the paper by Biswas et al. included only patients from a local state-wide registry who were readmitted within the first 30 days post-PCI, whereas our cohort is from a nationwide registry and included patients readmitted up to 1 year post-PCI, providing a more comprehensive understanding of the characteristics of patients readmitted and potentially enabling assessment of causes for unplanned readmission that may vary at different time-points following PCI.

Furthermore, there are several advantages for assessment from a national registry including large in size, well-resourced, robust governance structure, and more representative in comparison to regional registries. National registries may be less subject to biases introduced by regional practice trends, local socio-economic and cultural factors. Additionally, national registries may be more suited to informing governments, health providers, researchers and health economists compared to regional registries. Although the study by Kwok et al. is large, it only assessed patients with readmission up to 180 days post-PCI.

Finally, neither the Biswas nor Kwok papers examined long-term outcomes after unplanned readmission post-PCI, likely as both studies draw from administrative databases that do not provide long-term follow-up. In contrast, our paper is based on data from a prospective long-term clinical registry that performs annual patient follow-up and used this to examine very long-term outcomes 3 years post-PCI of patients experiencing unplanned readmission within the first 30 days or 1 year of PCI.

As a result, we believe there is still a need to explore the predictors of unplanned readmission and their impact on long-term outcomes post-PCI, and that our analysis expands the current literature. For the first time, we compare predictors of both unplanned readmission and long-term outcomes within the same large patient cohort, which will inform clinicians and potentially provide opportunities for intervention to improve patient outcomes.

4) The methods sections should be more detailed, including eligibility criteria and definition of outcomes (e.g., MACE).

Response:

We agree that the points above could be further clarified.

Action:

We have expanded the methods section as recommended by the Reviewer. These changes can be found on pages 2-3.

Response:

We thank the reviewer for noting this important point.

Action:

We have added a clarification in the Results and Methods section to explain that of patients experiencing unplanned cardiac readmission, 10.9% who were planned for staged PCI were readmitted prematurely and unexpectedly due to a new event such as angina or acute coronary syndrome. Patients who were planned for staged PCI after their index procedure and who did not have a further acute clinical event were not included in the Unplanned Readmission group for this analysis. These changes can be found on page 3, line no 109-117 and page 3 line 143-147.

Response:

We agree with the reviewer that the manuscript could be improved by more clearly summarising the analyses of unplanned readmission predictors, which are presented in the results section, within the conclusions and abstract. We have revised these sections accordingly.

Author Response File: Author Response.docx

Reviewer 2 Report

In the current study, the authors evaluate the effect of readmission after PCI at 30 days, 1 year and 3 years and try to identify predictors of worse outcomes.

Major comments:

The major findings from the study is probably something that most practitioners already take for granted, readmissions after PCI increases risk, major predictors of worse outcomes are high risk markers such as CKD, CHF, complex lesions, bleeding etc. It is good to see that published from a large registry like the one presented yet we can probably obtain more useful information from a better and more detailed analysis of this cohort of patients.

The authors tackle the issue of private vs public hospitals hence they need to provide data from their cohort on what percentage of the patients in the registry were from the private vs public hospitals and what is the percentage readmission between the different cohorts and is there difference in outcomes? There maybe some difference in baseline characteristics of patients seeking care in private vs public hospitals too.

The authors need to stratify the outcomes based on whether the PCI was done for stable disease or ACS?

Half of the readmissions were due to recurrent chest pain. This is a very important point as a lot of patients with stable CAD can actually have non cardiac chest pain and hence their pain never really improves after PCI. I would like to see the data on the patients who were readmitted with chest pain and were ruled out whether it confirms risk or not.

Minor comments:

In the introduction: “Unplanned readmissions may reflect incomplete or lower quality hospital care, poor coordination of services, post-discharge or an adverse outcome from complications such as bleeding from dual 55 anti-platelet therapy that have been demonstrated to increase mortality after PCI”.

Readmissions maybe related to the fact that patients with coronary artery disease requiring revascularizations have a lot of other comorbidities and can get readmitted whether they undergo PCI or not, so I would rephrase the sentence better using more cautious language.

Please define MACE in your study.

Author Response

Reviewer #2 Comments:

We thank the reviewer for their comments and thoughtful review.

Comments and Suggestions for Authors

In the current study, the authors evaluate the effect of readmission after PCI at 30 days, 1 year and 3 years and try to identify predictors of worse outcomes.

Major comments:

Response:

We agree that a comparison between patients treated in public and private settings would be of interest. However, for this cohort study data on only approximately 600 public patients was available, which is insufficient with the existing stratification by the timing of readmission to permit a useful comparative analysis with private patients.

Further, we have published data indicating that the private hospital patients in the Australian GCOR PCI registry have similar baseline characteristics to public hospital patients such as in the Victorian MIG and VCOR PCI registries, so we would not expect to find major differences between public and private patients enrolled in GCOR. (Reference 14 Heart Lung and Circulation(2017)26; 1303-1309)

The authors need to stratify the outcomes based on whether the PCI was done for stable disease or ACS?

Response:

We have controlled for the variable of ACS vs elective PCI in the propensity-matched sampling, and in the multivariate regression analysis for Early vs Late unplanned readmission. An additional stratification of the data by ACS vs elective would leave insufficient patient numbers in some groups for meaningful analysis. A separate analysis of Unplanned readmissions by Acute vs Elective status was beyond the scope of this paper, however, we agree this would be a good topic for a separate analysis and manuscript in the future.

Response:

The reviewer has identified an important point regarding potential causes of unplanned readmission. Unfortunately, we do not have access to hospital-level data regarding individual patient investigations such as ECGs, pathology and non-invasive imaging that would be required to assess each patient’s chest pain as cardiac or non-cardiac in origin, as only discharge summary data is provided to the registry. An analysis such as that proposed would we agree would be informative however is beyond the scope of this paper

Minor comments:

In the introduction: “Unplanned readmissions may reflect incomplete or lower quality hospital care, poor coordination of services, post-discharge or an adverse outcome from complications such as bleeding from dual anti-platelet therapy that have been demonstrated to increase mortality after PCI”.

Response:

We agree with the reviewer’s comment that patients with coronary artery disease are at high risk of further cardiovascular events such as readmission. However, we note the comment regarding bleeding and readmission was in reference to the analysis by Kwok of a large dataset, and feel it is pertinent background to our analysis (Circ Cardiovasc Interv. 2015 Apr;8(4):e001645.)

Please define MACE in your study.

Action:

We have expanded the methods section as recommended by the Reviewer. These changes can be found on page 3, line no.131.