Incidence, trends, and outcomes of infection sites among hospitalizations of sepsis: A nationwide study.

PURPOSE: To determine the trends of infection sites and outcome of sepsis using a national population-based database.
MATERIALS AND METHODS: Using the Nationwide Inpatient Sample database of the US, adult sepsis hospitalizations and infection sites were identified using a validated approach that selects admissions with explicit ICD-9-CM codes for sepsis and diagnosis/procedure codes for acute organ dysfunctions. The primary outcome was the trend of incidence and in-hospital mortality of specific infection sites in sepsis patients. The secondary outcome was the impact of specific infection sites on in-hospital mortality.
RESULTS: During the 9-year period, we identified 7,860,687 admissions of adult sepsis. Genitourinary tract infection (36.7%), lower respiratory tract infection (36.6%), and systemic fungal infection (9.2%) were the leading three sites of infection in patients with sepsis. Intra-abdominal infection (30.7%), lower respiratory tract infection (27.7%), and biliary tract infection (25.5%) were associated with highest mortality rate. The incidences of all sites of infections were trending upward. Musculoskeletal infection (annual increase: 34.2%) and skin and skin structure infection (annual increase: 23.0%) had the steepest increase. Mortality from all sites of infection has decreased significantly (trend p<0.001). Skin and skin structure infection had the fastest declining rate (annual decrease: 5.5%) followed by primary bacteremia (annual decrease: 5.3%) and catheter related bloodstream infection (annual decrease: 4.8%).
CONCLUSIONS: The anatomic site of infection does have a differential impact on the mortality of septic patients. Intra-abdominal infection, lower respiratory tract infection, and biliary tract infection are associated with higher mortality in septic patients.

Introduction

Being one of the most expensive conditions to treat and a leading cause of death, sepsis has become a major health problem [1, 2]. The incidence of sepsis has been steadily increasing in the past decade, and one recent study estimated an increase in sepsis admissions from 143,000 in 2000 to 343,000 in 2007 [3]. Sepsis was ranked in the top four most costly conditions, costing an aggregate of $20,298,000 million yearly, in US hospitals between all four payer groups (Medicare, Medicaid, private insurers, and uninsured) [4]. This burden on the healthcare system has led to researchers attempting to redefine sepsis and understand its pathophysiologic basis [5, 6]. A recent taskforce led by the Society of Critical Care Medicine and the European Society of Intensive Care Medicine convened and redefined sepsis as life threatening organ dysfunction caused by a dysregulated host response to infection [7]. Current knowledge suggests that mortality in sepsis is related to an overwhelming host immune response to invading pathogens infecting a specific anatomic site, and in current practice the suspected site of infection dictates treatment decisions that impact patient outcome [8]. Therefore, it’s probable that the anatomic infection site may have a significant impact on sepsis mortality. However, there has been a paucity of studies with inconsistent results addressing the various infectious sites effects on mortality, and no reports on the temporal trends of infectious sites and their outcomes [9-12]. Another aspect that could be influenced by studying current trends of infectious sites and their outcome could be researching specific preventative measures tailored towards the most common or highest risk infectious site. Current interventions to prevent certain anatomic site infections are in place such are vaccination against pneumococcal pneumonia or ventilator and line associated bundles [13, 14]. Thus, a study directed towards investigating these issues is important for intensive care resource allocation, public health prevention, and helping prioritize future research.

The primary aim of this study was to delineate the change in the incidence and in-hospital mortality of specific infection sites in sepsis patients over time. The secondary aim was to investigate the effect of anatomic infection site on the in-hospital mortality of sepsis patients.

Methods

Data sources

This study was conducted using 2006–2014 data from the Nationwide Inpatient Sample (NIS), part of the Healthcare Cost and Utilization Project, a federal-state-industry partnership sponsored by the Agency for Healthcare Research and Quality (AHRQ). The NIS is the largest all-payer inpatient database in the US, which is a 20% stratified sample of all US community hospitals as defined by the American Hospital Association: nonfederal, short-term, general, and specialty hospitals whose facilities are open to the public. By weighting the patient-level discharge data, it estimates more than 35 million hospitalizations nationally. The database includes clinical variables on all diagnoses and procedures occurring during each hospital admissions. Since the NIS database contains de-identified information regarding each hospitalization, the need for informed consent was waived [15].

Case selection and definitions

Sepsis hospitalizations were identified using a validated approach that selects admissions with relevant International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) diagnosis/procedure codes. Conforming to Sepsis-3 definition, sepsis is defined as life-threatening organ dysfunction caused by a dysregulated host response to infection. The coding system proposed and validated previously by Martin GS et al. is a more conservative estimates that showed a parallel trend with the electronic health record (EHR) estimates [16, 17]. Therefore, we used the Martin’s criteria to identify patients with sepsis in this study. (S2 Table) Sensitivity analysis using Angus criteria was performed to corroborate the results. Operationally, we identified cases with sepsis by selecting all cases with explicit ICD-9-CM codes for sepsis or systemic fungal infection (038 septicemia, 020.0 septicemic, 790.7 bacteremia, 117.9 disseminated fungal infection, 112.5 disseminated candida infection, or 112.81 (disseminated fungal endocarditis) and a diagnosis of acute organ dysfunction. Site of infection was categorized as lower respiratory tract infection, genitourinary tract infection, skin and skin structure infection, catheter related bloodstream infection, intra-abdominal infection, systemic fungal infection, primary bacteremia, musculoskeletal infection, and biliary tract infection (S3 Table). Acute organs/systems dysfunction used for this study was: cardiovascular, respiratory, central nervous system, hematologic, hepatic, renal and metabolic system dysfunction. Shock was included as a form of cardiovascular dysfunction. For patient with multiple diagnoses, only primary and secondary diagnoses were recorded. We used Elixhauser comorbidity Index as our comorbidity index. The following information was collected for analysis: demographic, presence of pre-existing comorbidity, and outcome.

Outcome measures

The primary outcome was the trend of incidence and in-hospital mortality of specific infection sites in sepsis patients. The secondary outcome was the impact of specific infection sites on in-hospital mortality.

Statistical analyses

Data management and statistical analyses were conducted using SAS (SAS Inc, Cary, NC) and SAS-callable SUDAAN software (version 9.4, RTI International, Research Triangle, NC) to account for the stratified sampling design used to collect the hospital discharge data. The frequency of hospitalizations for sepsis with specific type of infection was estimated following recommendations from the AHRQ. By using survey-specific statements, SURVEYMEANS in SAS program, we weighted the patient-level discharge data using the weights provided in the NIS database. Continuous variables with normal distribution were presented as mean with standard error (SE), and non-normal variables were reported as median with interquartile range (IQR). Categorical variables were reported as percentage (%). We calculated the overall and average annual percent change in the hospitalization and mortality of sepsis and specific site of infection between 2006 and 2014. To examine the significance of trends of incidence and mortality, we performed linear regression analysis. To evaluate the impact of individual site of infection on the survival of sepsis patients, we fit a multivariable logistic regression model adjusting for age, sex, and comorbidity measures. We used the entire study period (2006 through 2014) for this regression analysis to ensure adequate power to make reliable estimates of risk. Because the mortality rate for patients with sepsis is higher than 10% in this analysis, the rare disease assumption does not hold. As a result, risk ratios cannot be estimated by odds ratios. We used the formula proposed by Zhang and Yu to approximate the relative risk [18]. Two-sided P <0.01 was considered statistically significant for all analyses.

Results

During the 9 year period between 2006 and 2014, we identified 7,860,687 admissions of adult sepsis. Fig 1 shows the cohort assembling process, total number of each site of infection, corresponding mortality rate, and total number of deaths for each site of infection. Genitourinary tract infection, lower respiratory tract infection and systemic fungal infection were the leading three sites of infection in patients with sepsis, accounting for 36.70%, 36.55% and 9.22% of all sites of infection, respectively. Intra-abdominal infection, lower respiratory tract infection, and biliary tract infection were associated with poor outcome, with a mortality rate of 30.65%, 27.70%, and 25.48%, respectively. Primary bacteremia, musculoskeletal infection, and catheter-related bloodstream infection, however, were associated with better outcome, with a mortality rate of 7.43%, 14.14%, and 15.36%, respectively. Taking the incidence and mortality rate together, lower respiratory tract infection was the leading cause of mortality (weighted death number = 795,825), followed by genitourinary tract infection (weighted death number = 489,964) and systemic fungal infection (weighted death number = 153,027). Table 1 shows the characteristics and sites of infection in the three sub-periods. There are more male patients than female in all subperiods. The mean age of sepsis patients was comparable over the subperiods. The incidence of comorbidities in patients with sepsis increased over the three sub-periods.

Fig 1

Flowchart of patients in this study.

Table 1

Characteristics of study cohort, stratified by three periods between 2006 and 2014.

Characteristic	2006–2008n = 1,957,110	2009–2011n = 2,695,151	2012–2014n = 3,208,425
Age,yrs	68.22±0.18	67.96±0.17	67.61±0.06
Male sex, %	985407(50.35%)	1366596(50.71%)	1635850(50.99%)
Comorbidity
Combined comorbidity score	13.47±0.09	14.68±0.1	15.02±0.03
Hypertension	762695(38.97%)	1405676(52.16%)	1869545(58.27%)
Congestive heart failure	482731(24.67%)	654185(24.27%)	795325(24.79%)
Chronic pulmonary disease	444383(22.71%)	657182(24.38%)	848340(26.44%)
Chronic renal failure	488108(24.94%)	746443(27.7%)	928730(28.95%)
Uncomplicated diabetes	329425(16.83%)	628498(23.32%)	818090(25.5%)
Coagulopathy	347156(17.74%)	543134(20.15%)	677745(21.12%)
Neurological disorders	226111(11.55%)	390633(14.49%)	509680(15.89%)
Weight loss	294112(15.03%)	575848(21.37%)	658120(20.51%)
Valvular heart disease	113762(5.81%)	164232(6.09%)	228370(7.12%)
Diabetes with complications	114315(5.84%)	213520(7.92%)	299655(9.34%)
Depression	83354(4.26%)	240851(8.94%)	351695(10.96%)
Peripheral vascular disease	104608(5.35%)	243010(9.02%)	323020(10.07%)
Chronic liver disease	94970(4.85%)	157011(5.83%)	219140(6.83%)
Obesity	68749(3.51%)	262865(9.75%)	451700(14.08%)
Alcohol abuse	75366(3.85%)	120961(4.49%)	169145(5.27%)
Metastatic cancer	101816(5.2%)	137989(5.12%)	166030(5.17%)
Paralysis	96977(4.96%)	184049(6.83%)	222930(6.95%)
Psychoses	61897(3.16%)	134534(4.99%)	185210(5.77%)
Solid tumor	66646(3.41%)	104102(3.86%)	131915(4.11%)
Rheumatic disease	42974(2.2%)	87818(3.26%)	120520(3.76%)
Drug abuse	38803(1.98%)	67415(2.5%)	115210(3.59%)
Lymphoma	42402(2.17%)	57161(2.12%)	66465(2.07%)
AIDS	27871(1.42%)	31230(1.16%)	30165(0.94%)
Sites of infection
Lower respiratory tract infection	700727(35.8%)	1000502(37.12%)	1171990(36.53%)
Genitourinary tract infection	689089(35.21%)	1010380(37.49%)	1185475(36.95%)
Skin and skin structure infection	133718(6.83%)	218931(8.12%)	285830(8.91%)
Catheter related bloodstream infection	144494(7.38%)	126390(4.69%)	130240(4.06%)
Intra-abdominal infection	95446(4.88%)	147403(5.47%)	175310(5.46%)
Biliary tract infection	11312(0.58%)	18168(0.67%)	22035(0.69%)
Systemic fungal infection	155501(7.95%)	283583(10.52%)	285330(8.89%)
Primary bacteremia	161846(8.27%)	201650(7.48%)	183280(5.71%)
Musculoskeletal infection	42911(2.19%)	83271(3.09%)	105455(3.29%)

n = total episodes of sepsis hospitalization in the subperiod; values are n, mean ± SE, or n (%)

Fig 2 paired with Table 2 shows the changes in population incidence of specific site of infection in patients with sepsis. The incidence of all sites of infections were trending upward. Musculoskeletal infection, skin and skin structure infection and biliary tract infection had the steepest increase, with an annual increase rate of 34.22%, 23.02% and 20.07%, respectively. On the contrary, Catheter related bloodstream infection and primary bacteremia had a decrease or slow increase, with an annual change rate of -0.97% and 2.89%, respectively. Other sites of infection had an annual increase rate between 13.67% and 18.94%. The aforementioned temporal changes in incidence were all significant (Trend p value <0.001).

Fig 2

Changes in number of sepsis hospitalizations by specific infection sites among patients with sepsis, from 2006 to 2014.

(A) High to moderate number of hospitalizations, (B) low number of hospitalizations.

Table 2

Weighted number of sepsis hospitalizations by specific infection site among patients with sepsis.

The annual incidence is presented by events per 100,000 hospitalizations.

	2006	2010	2014	Annual change, %
Lower respiratory tract infection	1.94	3.31	4.32	13.67%
Genitourinary tract infection	1.89	3.29	4.35	14.43%
Intra-abdominal infection	0.26	0.50	0.65	16.15%
Skin and skin structure infection	0.36	0.72	1.11	23.02%
Musculoskeletal infection	0.10	0.27	0.41	34.22%
Primary bacteremia	0.49	0.68	0.61	2.89%
Catheter related bloodstream infection	0.52	0.42	0.47	-0.97%
Systemic fungal infection	0.37	0.93	1.00	18.94%
Biliary tract infection	0.03	0.06	0.08	20.07%

Fig 3 paired with Table 3 shows the temporal trends of mortality rate for each infection site in patients with sepsis. Mortality from all sites of infection has decreased significantly in the study period (trend p<0.001). Skin and skin structure had the fastest declining rate (annual decrease: 5.51%) followed by primary bacteremia (annual decrease: 5.32%) and catheter related bloodstream infection (annual decrease: 4.82%).

Fig 3

Temporal trend of mortality rate for specific site of infections among patients with sepsis.

Table 3

In-hospital mortality rate and annual change in rate for specific infection site among patients with sepsis.

	2006	2010	2014	Annual change, %
Lower respiratory tract infection	34.99%	28.46%	23.07%	-3.79%
Genitourinary tract infection	22.70%	17.05%	13.41%	-4.55%
Intra-abdominal infection	37.10%	31.56%	26.18%	-3.27%
Skin and skin structure infection	23.01%	15.99%	11.60%	-5.51%
Musculoskeletal infection	18.55%	14.60%	11.54%	-4.20%
Primary bacteremia	10.98%	7.04%	5.72%	-5.32%
Catheter related bloodstream infection	20.28%	14.63%	11.49%	-4.82%
Systemic fungal infection	26.71%	21.49%	18.18%	-3.55%
Biliary tract infection	30.30%	24.91%	22.08%	-3.01%

Fig 4 shows the adjusted relative risk with 95% confidence intervals of infection site on the outcome of sepsis. Using primary bacteremia as reference, sepsis patients with intra-abdominal infection had the highest mortality (RR:4.21), followed by lower respiratory tract infection (RR: 3.84), biliary tract infection (RR: 3.24), systemic fungal infection (RR: 2.77), skin and skin structure infection (RR: 2.29), musculoskeletal infection (RR: 2.27), genitourinary traction infection (RR:2.19), or catheter related bloodstream infection (RR:2.15). Sensitivity analysis with Angus criteria showed similar trend of sepsis as our main results (S4–S7 Tables and S1–S3 Figs)

Fig 4

Survival impact of individual infection site in relation to primary bacteremia.

The risk estimates were adjusted for all covariates listed in supporting S1 Table. RR refers to the relative risk. LCL and UCL refer to lower and upper confidence limits, respectively.

Discussion

Based on our study, there has been an increasing trend in the incidence of hospitalizations from sepsis with the greatest number of hospitalizations from lower respiratory tract infections and the least from biliary tract infections. There was also a trend of decreasing mortality from sepsis. Zahar et al suggested that neither site of infection nor presence of bacteremia associated with mortality [9]. However, that study was a single center study with limited sample size. Our study used national database to expand sample size and increase statistic power. Our study showed that, independent of predisposing factors, the site of infection is associated with in hospital mortality in patients with sepsis. Hospital mortality was highest for patients with intra-abdominal infection and lowest for primary bacteremia. This study is the first large national cohort study to investigate a relationship between site of infection and mortality. A few related studies have been performed using smaller sample sizes or different sepsis definitions. They found that either urosepsis or skin infections have a more favorable prognosis while pneumonia or intra-abdominal infection have worse prognosis. Multiple prior studies are consistent with our finding of the trend of increasing sepsis incidence with decreasing mortality [1–3, 19]. This trend is presumably reflecting ongoing efforts to improve sepsis awareness, treatment, documentation, and coding. For example, the surviving sepsis campaign started at the beginning of this study could account for early sepsis recognition and decreasing mortality with early antibiotic administration and three-hour bundle therapy [20-22].

The infection site with the highest incidence in this study was lower respiratory tract infections. Currently preventative strategies mainly aimed at streptococcus pneumoniae, which is the common pathogen of pneumonia. After the introduction of pneumococcal vaccinations to both pediatric and adult populations, incidence of pneumococcal pneumonia has decreased [23], A similar strategy of developing further vaccinations or improving upon current vaccinations against other pathogens could minimize predisposition to bacterial pneumonia.

Our study found that sites (intra-abdominal, respiratory, and biliary infections) that have the potential to develop a high burden of organisms resulting in a large downstream pro-inflammatory state caused the highest mortality [8, 24, 25]. In contrast, those infections with multiple protective barriers (such as skin, and musculoskeletal infections) had lower mortality rates [26]. This knowledge could be used to refine prognostication in sepsis helping to select patient populations that may benefit from novel treatments or that require higher levels of monitoring [27-29]. For example, it has been postulated that immunomodulatory agents failed to improve outcomes in septic patients in clinical trials because of enrolment of patients who have lower/intermediate risks or death [30]. Focusing treatments like these on higher mortality sites of infection could have an impact on these infectious sites. The differing mortalities in sites of infection could also be considered in choosing when antibiotic de-escalation is appropriate. Those patient with lower risk of mortality might benefit from early de-escalation of the antibiotics. Helping to reduce antimicrobial resistance, and adverse drug reactions [31, 32].

There were several limitations of this study. First, identifying sepsis using ICD-9 CM codes algorithm may not be as precise as screening EHR with clinical criteria because clinicians and hospital coders may vary widely in their knowledge and application of sepsis definitions [33]. However, the estimation of sepsis trend using EHR from several hospitals demands a lot of resources. In addition, different hospitals contributed data of different years with different case mix lowers the generalizability of the estimation. Previous studies showed sepsis estimated from Martin’s algorithm is a conservative and reasonable proxy to the estimates from EHR, therefore we adopted Martin’s implementation for this study [17, 34]. Second, although we adjusted for multiple factors that could influence hospital mortality, there may be other confounding factors that we did not account for and measure. Third, we assumed organ dysfunction to be a downstream effect of serious infection, and thus did not adjust for organ dysfunction in the regression model avoiding intermediate bias. Fourth, by using in-hospital mortality as our endpoint overall mortality of specific infections may have been underestimated if the events occurred outside of the hospital. Moreover, our administrative data is unable to establish a firm temporal relationship between sepsis and the onset of organ dysfunction. Meanwhile, due to the insufficient information from the database, we cannot identify patients’ socioeconomical status, community-acquired or nosocomial, medical or surgical hospitalization. Also, we didn’t perform the control group analysis due to insufficient data. Further study involving more detailed in-patient data is required. In order to increase the comparability of our study, we used Angus implementation for the recognition of infection sites. The Angus implementation was originally invented for the identification of severe sepsis with an ICD-9 based criteria, which is by far one of the most widely used implementations [9, 34–38]. Using the same criteria as other studies could increase the comparability and provide opportunity for future meta-analysis. However, some of the specific diagnoses may not be included. Lastly, our results may not be generalizable to other parts of the world because this study was conducted in American hospitals. Further studies would be needed to address these limitations and provide explanation to this trend.

Conclusion

There is a significant difference in the trend of incidence and outcome of sepsis from different anatomic sites of infection. Clinicians should be aware of different anatomic sites of infection could cause higher mortality in septic patients such as intra-abdominal infection, lower respiratory tract infection, and biliary tract infection.

Sensitivity Test—Changes in number of sepsis hospitalizations by specific infection sites among patients with sepsis, from 2006 to 2014.

(TIFF)

Click here for additional data file.

Sensitivity Test—Temporal trend of mortality rate for specific source of infections among patients with sepsis.

(TIFF)

Click here for additional data file.

Sensitivity Test—Survival impact of individual infection site in relation to primary bacteremia.

The risk estimates were adjusted for all covariates listed in Supporting S7 Table. RR refers to the relative risk. LCL and UCL refer to lower and upper confidence limits, respectively.

(TIFF)

Click here for additional data file.

Covariates with associated relative risk in the outcome regression model.

(PDF)

Click here for additional data file.

ICD-9 Code associated with organ dysfunction.

(PDF)

Click here for additional data file.

ICD-9-CM codes of site of infections associated with sepsis.

(PDF)

Click here for additional data file.

Sensitivity Test—Characteristics of study cohort, stratified by three periods between 2006 and 2014.

(PDF)

Click here for additional data file.

Sensitivity Test—Number of sepsis hospitalizations by specific infection site among patients with sepsis.

(PDF)

Click here for additional data file.

Sensitivity Test—In-hospital mortality rate and annual change in rate for specific infection site among patients with sepsis.

(PDF)

Click here for additional data file.

Sensitivity Test—Covariates with associated relative risk in the outcome regression model.

(PDF)

Click here for additional data file.

25 Aug 2019

PONE-D-19-15923

Incidence, trends, and outcomes of infection sites among patients with sepsis: a nationwide study

PLOS ONE

Dear Dr Lee,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

The reviewers raised some concerns regarding the chosen sepsis definition, particularly as up to 30% of patients with sepsis do not have positive cultures and therefore would not be captured in the current analysis. Therefore, sensitivity analysis using different sepsis definition criteria (implicit Angus criteria vs. explicit sepsis criteria) should be performed to corroborate the presented results. In addition, ICD-9 codes used to define 'organ dysfunction' should be listed in the supplement. Second, it is unclear whether the analyses included both community-acquired sepsis ('present on admission') or nosocomial sepsis. These are very different entities with regards to epidemiology, source of infection, risk of death, etc. Along these lines it would be interesting to know what proportion of hospitalizations were primary 'medical' or 'surgical'. Third, previous studies by Lindenauer (JAMA 2012) and Rhee (JAMA 2017) suggest that temporal trends in pneumonia and sepsis estimates are associated with differences in coding - additional sensitivity analysis should be performed to verify or refute these findings. An additional limitation is the inability of administrative data to establish a firm temporal relationship between sepsis and the onset of organ dysfunction. This important limitation should be added to the discussion section. Finally, the discussion should be expanded to contrast the findings to preexisting literature. For example,  Zahar JR et al previously reported no association between infection site and presence of bacteremia with mortality.

We would appreciate receiving your revised manuscript by Oct 09 2019 11:59PM. When you are ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter.

To enhance the reproducibility of your results, we recommend that if applicable you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). This letter should be uploaded as separate file and labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. This file should be uploaded as separate file and labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. This file should be uploaded as separate file and labeled 'Manuscript'.

Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.

We look forward to receiving your revised manuscript.

Kind regards,

Florian B. Mayr

Academic Editor

PLOS ONE

Journal Requirements:

1. When submitting your revision, we need you to address these additional requirements.

Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

http://www.journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and http://www.journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. Please include your tables as part of your main manuscript and remove the individual files. Please note that supplementary tables (should remain/ be uploaded) as separate "supporting information" files

3. Thank you for stating the following in the Acknowledgments Section of your manuscript:

This work is partly supported by NTUH.107-P03 grant

a) We note that you have provided funding information that is not currently declared in your Funding Statement. However, funding information should not appear in the Acknowledgments section or other areas of your manuscript. We will only publish funding information present in the Funding Statement section of the online submission form.

b) Please provide an amended statement that declares *all* the funding or sources of support (whether external or internal to your organization) received during this study, as detailed online in our guide for authors at http://journals.plos.org/plosone/s/submit-now.  Please also include the statement “There was no additional external funding received for this study.” in your updated Funding Statement.

Please remove any funding-related text from the manuscript and let us know how you would like to update your Funding Statement. Currently, your Funding Statement reads as follows:

The author(s) received no specific funding for this work.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Partly

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: Mann et al. have conducted a longitudinal analysis of sepsis outcomes by site of infection, over an extended period, using the National Inpatient Sample database, representing a large portion of the U.S. population. The authors have performed an appropriate analysis based on the data available and provided trends in mortality by infectious site over time, an important finding. However, I have some reservations about the methodology used provided in comments below.

Major revisions:

1. Methods: The authors’ primary conclusion in this study is that site of infection affects mortality and is changing over time. However, identification of site of infection seems inconsistent.

a. (Supplemental Table 2) Certain infections appear to be omitted from the inclusion diagnoses (i.e. cholangitis, Clostridium dificile colitis, CNS infections, endocarditis, and GU infections in women such as endometritis or ovarian abscess). Suggest improving identification of source of infection as this is the most important variable in the analysis.

b. Second, how do the authors identify the primary site of infection? For example, if a hospitalization is coded for intestinal perforation and bacteremia and fungemia, how would this hospitalization be grouped in their analysis? Would it be the highest ranking diagnosis code or would it be included in all groups. This is important for assumptions of mortality in each group. It appears that these groups are mutually exclusive but it is not clear in the methods section.

c. Does primary bacteremia mean no other diagnosis could be found for site of infection? More detail is required since there is a significant decrease in this site of infection over time. I suspect the outcomes and numbers of this site of infection would change greatly once item 1a is addressed.

2. Methods Line 95. The authors identify sepsis using methods by Martin et al. There are several other methods for identifying sepsis including Dombrovskiy et al. (PMID: 17414736). To improve the confidence of the conclusions, a sensitivity analysis using one of the other methods would be helpful.

3. Methods Line 126. The adjusted analysis includes age, comorbidities, and gender. Are there other SES variables in NIS, such as SES, income, insurance type, that should be included in the adjustment? Agree that organ dysfunction should not be included in the regression model.

Minor revisions:

1. Title Line 1. Suggest changing patients to hospitalizations. The NIS includes hospitalizations in which one patient could be readmitted. Therefore, each hospitalization may not be unique to one patient.

2. Abstract Line 26. There are places (lines 27 and 37) where site of infection and source of infection are used interchangeably. Suggest using one consistent word such as site for all subsequent descriptions.

3. Introduction Line 49. “due to being one” sounds confusing. Suggest re-wording.

4. Intro Line 52. Suggest changing “Septicemia” to Sepsis.

5. Methods Line 78. Please provide rationale for study period 2006-2014. Is it because sepsis codes were developed in 2003 and there was late adoption? Also provide rationale for the break points chosen in Tables 1, 2, and 3.

6. Methods Line 94. Shouldn’t the acronym be EHR?

7. Methods Line 131. Consider making significance level <0.01 given size of sample (see PMID:30398593).

8. Results line 143-144. Is death number the weighted estimate of in-hospital mortality or the raw number from NIS? Would suggest rewording death number.

9. Results Line 163. Change intra-abdomen to intra-abdominal

10. Discussion Lines175-176. Reword septic source

11. Discussion Line 177. Which studies? Please include references

12. Discussion Lines 193-194. I’m not sure the urinary system has more protective barriers than the respiratory system. This sentence sounds strange. Would re-word.

13. Discussion Line 193. Remove period after mortality

14. Discussion Line 203. Suggest re-writing. What does tolerate mean?

15. Discussion Line 211. Change HER to EHR

16. Conclusion Lines 220-225. This feels vague. What specific trend do you want readers to take away and what specific future advancement does this study help with?

17. Tables 1, 2, and 3. Are these all weighted numbers? Please clarify. Suggest changing to title to reflect weighted analysis

18. Table 1. What comorbidity score was chosen? Please reference in methods.

19. Table 2. Change “systematic” to systemic. Double check throughout.

20. Figure 1. Change “miss” to missing values

21. Figure 2b. Y axis title is not fully seen

22. Figures. Suggest placing legends underneath graphs

23. Figure 4. I like this.

24. Supplementary Table 1. Why was primary bacteremia chosen as the reference? Primary bacteremia without a source seems like a rare entity and as such should not be the comparison.

25. Supplementary Table 1. Almost all variables reach statistical significance due to the sample size. Is there a comparison population you could track over the time period (i.e. those requiring mechanical ventilation) to confidently say the sepsis trend is changing rather than an artifact of the large sample size? This could be a “control group analysis” throughout the methods, results, and discussion.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: Yes: Matthew K. Hensley

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files to be viewed.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at figures@plos.org. Please note that Supporting Information files do not need this step.

12 Oct 2019

Ref: PONE-D-19-15923

Title: Incidence, trends, and outcomes of infection sites among hospitalizations of sepsis: a nationwide study

Journal: PLOS ONE

Dear editor and reviewers:

Thank you for your e-mail with the referees’ comments. We have revised the manuscript as advised. The detailed point-by-point responses are as follows. We hope the manuscript is now acceptable to PLOS ONE.

Editor’s Comment:

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

The reviewers raised some concerns regarding the chosen sepsis definition, particularly as up to 30% of patients with sepsis do not have positive cultures and therefore would not be captured in the current analysis. Therefore, sensitivity analysis using different sepsis definition criteria (implicit Angus criteria vs. explicit sepsis criteria) should be performed to corroborate the presented results.

Author Reply:

Thanks for your comment. We have added our sensitivity analysis using Angus criteria into the supplementary section accordingly and added relevant description into our manuscript. Similar to our analysis, sensitivity analysis demonstrated increasing trend of sepsis. Meanwhile, the sensitivity test yielded similar result as our main analysis.

“Sensitivity analysis using Angus criteria was performed to corroborate the result.” (Section of Method, line 99)

“Sensitivity analysis with Angus criteria showed similar trend of sepsis as our main results. (S4-S7 Table and S1-3 Figs)” (Section of Results, line 187-189)

In addition, ICD-9 codes used to define 'organ dysfunction' should be listed in the supplement.

Author Reply:

Thanks for your advice. We have added into our manuscript as supporting table 2.

Second, it is unclear whether the analyses included both community-acquired sepsis ('present on admission') or nosocomial sepsis. These are very different entities with regards to epidemiology, source of infection, risk of death, etc.

Author Reply:

Thanks for your advice. We agree that community-acquired and nosocomial sepsis have distinctive differences. However, due to the limitation of the database, we didn’t have sufficient data to determine whether it is community-acquired or nosocomial. We have added to our limitation section.

“Meanwhile, due to the insufficient information from the database, we cannot identify patients’ socioeconomical status, community-acquired or nosocomial, medical or surgical hospitalization. Also, we didn’t perform the control group analysis due to insufficient data. Further study involving more detailed in-patient data is required.” (Section of Discussion, line 246-249)

Along these lines it would be interesting to know what proportion of hospitalizations were primary 'medical' or 'surgical'.

Author Reply:

Thanks for your advice. Due to the limitation of the database, we don’t have sufficient data to determine whether it is surgical or medical. We have added to our limitation section.

“Meanwhile, due to the insufficient information from the database, we cannot identify patients’ socioeconomical status, community-acquired or nosocomial, medical or surgical hospitalization. Also, we didn’t perform the control group analysis due to insufficient data. Further study involving more detailed in-patient data is required.” (Section of Discussion, line 246-249)

Third, previous studies by Lindenauer (JAMA 2012) and Rhee (JAMA 2017) suggest that temporal trends in pneumonia and sepsis estimates are associated with differences in coding - additional sensitivity analysis should be performed to verify or refute these findings.

Author Reply:

Thank you for your comment. We have done a sensitivity test using Angus criteria to verify our results. We have revised the manuscript and added the sensitivity test results in the supplementary section accordingly.

“Sensitivity analysis using Angus criteria was performed to corroborate the result.” (Section of Method, line 99)

“Sensitivity analysis with Angus criteria showed similar trend of sepsis as our main results. (S4-S7 Table and S1-3 Figs)” (Section of Results, line 187-189)

An additional limitation is the inability of administrative data to establish a firm temporal relationship between sepsis and the onset of organ dysfunction. This important limitation should be added to the discussion section.

Author Reply:

Thanks for your advice. We have corrected the manuscript accordingly.

“Moreover, our administrative data is unable to establish a firm temporal relationship between sepsis and the onset of organ dysfunction.” (Section of Discussion, line 245-246)

Finally, the discussion should be expanded to contrast the findings to preexisting literature. For example, Zahar JR et al previously reported no association between infection site and presence of bacteremia with mortality.

Author Reply:

Thanks for your advice. We have revised the manuscript accordingly.

“Zahar et al suggested that neither site of infection nor presence of bacteremia associated with mortality. (9) However, that study was a single center study with limited sample size. Our study used national database to expand sample size and increase statistic power.” (Section of Discussion, line 197-198)

Journal Requirements:

1. When submitting your revision, we need you to address these additional requirements.

Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

http://www.journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and http://www.journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. Please include your tables as part of your main manuscript and remove the individual files. Please note that supplementary tables (should remain/ be uploaded) as separate "supporting information" files

3. Thank you for stating the following in the Acknowledgments Section of your manuscript:

This work is partly supported by NTUH.107-P03 grant

a) We note that you have provided funding information that is not currently declared in your Funding Statement. However, funding information should not appear in the Acknowledgments section or other areas of your manuscript. We will only publish funding information present in the Funding Statement section of the online submission form.

Author Reply:

Thanks for your advice. We didn’t receive any funding for this study. We have corrected the manuscript accordingly.

b) Please provide an amended statement that declares *all* the funding or sources of support (whether external or internal to your organization) received during this study, as detailed online in our guide for authors at http://journals.plos.org/plosone/s/submit-now. Please also include the statement “There was no additional external funding received for this study.” in your updated Funding Statement.

Please remove any funding-related text from the manuscript and let us know how you would like to update your Funding Statement. Currently, your Funding Statement reads as follows:

The author(s) received no specific funding for this work.

Author Reply:

Thanks for your advice. We didn’t receive any funding for this study. We have corrected the manuscript accordingly.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Partly

________________________________________

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

________________________________________

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

________________________________________

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

________________________________________

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: Mann et al. have conducted a longitudinal analysis of sepsis outcomes by site of infection, over an extended period, using the National Inpatient Sample database, representing a large portion of the U.S. population. The authors have performed an appropriate analysis based on the data available and provided trends in mortality by infectious site over time, an important finding. However, I have some reservations about the methodology used provided in comments below.

Major revisions:

1. Methods: The authors’ primary conclusion in this study is that site of infection affects mortality and is changing over time. However, identification of site of infection seems inconsistent.

a. (Supplemental Table 2) Certain infections appear to be omitted from the inclusion diagnoses (i.e. cholangitis, Clostridium dificile colitis, CNS infections, endocarditis, and GU infections in women such as endometritis or ovarian abscess). Suggest improving identification of source of infection as this is the most important variable in the analysis.

Author Reply:

Thank you for your comment. We used Martin’s criteria for the identification of sepsis. Because Martin’s criteria did not provide detailed definitions of different infection sites, we identified infection sites based on Angus ICD9-CM Sepsis Abstraction Criteria. (1)

Angus et al performed a large-scale, multicenter epidemiological study and implemented the identification of severe sepsis using an ICD-9 based algorithm. (1) The Angus implementation is by far one of the most well-known and highly cited implementations of an ICD-coded case definition for sepsis. (2-6) In addition, using the same criteria as other studies could increase the comparability of our study, and would provide opportunity for future systemic review and meta-analysis. We have included in our limitation.

“In order to increase the comparability of our study, we used Angus implementation for the recognition of infection sites. The Angus implementation was originally invented for the identification of severe sepsis with an ICD-9 based criteria, which is by far one of the most widely used implementations. (9, 34-38) Using the same criteria as other studies could increase the comparability and provide opportunity for future meta-analysis. However, some of the specific diagnoses may not be included.”

(Section of Discussion, line 250-255)

b. Second, how do the authors identify the primary site of infection? For example, if a hospitalization is coded for intestinal perforation and bacteremia and fungemia, how would this hospitalization be grouped in their analysis? Would it be the highest ranking diagnosis code or would it be included in all groups. This is important for assumptions of mortality in each group. It appears that these groups are mutually exclusive but it is not clear in the methods section.

Author Reply:

Thanks for your advice. Based on our method, we picked the top 2 (Primary and Secondary) diagnosis listed in the chart. The method has been clarified in the manuscript.

“For patient with multiple diagnoses, only primary and secondary diagnoses were recorded.” (Section of Method, line 109-110)

c. Does primary bacteremia mean no other diagnosis could be found for site of infection? More detail is required since there is a significant decrease in this site of infection over time. I suspect the outcomes and numbers of this site of infection would change greatly once item 1a is addressed.

Author Reply:

Thanks for your advice. We used the ICD-9 code 790.7 for the identification of primary bacteremia listed as primary diagnosis. No other diagnosis of infection sites was found.

2. Methods Line 95. The authors identify sepsis using methods by Martin et al. There are several other methods for identifying sepsis including Dombrovskiy et al. (PMID: 17414736). To improve the confidence of the conclusions, a sensitivity analysis using one of the other methods would be helpful.

Author Reply:

Thanks for your comment. We have added our sensitivity analysis using Angus criteria into the supplementary section accordingly and added relevant description into our manuscript. Similar to our analysis, sensitivity analysis demonstrated increasing trend of sepsis. Meanwhile, the sensitivity test yielded similar result as our main analysis.

“Sensitivity analysis using Angus criteria was performed to corroborate the result.” (Section of Method, line 99)

“Sensitivity analysis with Angus criteria showed similar trend of sepsis as our main results. (S4-S7 Table and S1-3 Figs)” (Section of Results, line 187-189)

3. Methods Line 126. The adjusted analysis includes age, comorbidities, and gender. Are there other SES variables in NIS, such as SES, income, insurance type, that should be included in the adjustment? Agree that organ dysfunction should not be included in the regression model.

Author Reply:

Thanks for your advice. Due to the limitation of the database, we didn’t have sufficient data to determine socioeconomical status. We have added to our limitation section as follows.

“Meanwhile, due to the insufficient information from the database, we cannot identify patients’ socioeconomical status, community-acquired or nosocomial, medical or surgical hospitalization. Also, we didn’t perform the control group analysis due to insufficient data. Further study involving more detailed in-patient data is required.” (Section of Discussion, line 246-249)

Minor revisions:

1. Title Line 1. Suggest changing patients to hospitalizations. The NIS includes hospitalizations in which one patient could be readmitted. Therefore, each hospitalization may not be unique to one patient.

Author Reply:

Thanks for your advice. We have corrected the manuscript accordingly. The title would be changed to “Incidence, Trends, and Outcomes of Infection Sites among Hospitalizations of Sepsis: a Nationwide Study.”

2. Abstract Line 26. There are places (lines 27 and 37) where site of infection and source of infection are used interchangeably. Suggest using one consistent word such as site for all subsequent descriptions.

Author Reply:

Thanks for your advice. We have corrected the manuscript accordingly. We have replaced all “source of infection” to “site of infection.”

3. Introduction Line 49. “due to being one” sounds confusing. Suggest re-wording.

Author Reply:

Thanks for your advice. We have corrected the manuscript accordingly.

“Being one of the most expensive conditions to treat and a leading cause of death, sepsis has become a major health problem.” (Section of Introduction, line 53-54)

4. Intro Line 52. Suggest changing “Septicemia” to Sepsis.

Author Reply:

Thanks for your advice. We have corrected the manuscript accordingly. (Section of Introduction, line 56)

5. Methods Line 78. Please provide rationale for study period 2006-2014. Is it because sepsis codes were developed in 2003 and there was late adoption? Also provide rationale for the break points chosen in Tables 1, 2, and 3.

Author Reply:

Thanks for your advice. Because ICD-10 was initiated since 2015, we choose a 9-year study period between 2006 and 2014 in our study. In consideration of the readability, we divided our population into 3 groups with same period of time for further comparison.

6. Methods Line 94. Shouldn’t the acronym be EHR?

Author Reply:

Thanks for your advice. We have corrected the manuscript accordingly.

7. Methods Line 131. Consider making significance level <0.01 given size of sample (see PMID:30398593).

Author Reply:

Thanks for your advice. We have corrected the manuscript accordingly.

8. Results line 143-144. Is death number the weighted estimate of in-hospital mortality or the raw number from NIS? Would suggest rewording death number.

Author Reply:

Thanks for your advice. The death number is the weighted number. We have clarified in our manuscript.

“Taking the incidence and mortality rate together, lower respiratory tract infection was the leading cause of mortality (weighted death number=795,825), followed by genitourinary tract infection (weighted death number=489,964) and systemic fungal infection (weighted death number=153,027). (Section of Results, line 147-150)

9. Results Line 163. Change intra-abdomen to intra-abdominal

Author Reply:

Thanks for your advice. We have corrected the manuscript accordingly.

10. Discussion Lines175-176. Reword septic source

Author Reply:

Thanks for your advice. We have corrected the manuscript accordingly. We have reworded it to “site of infection.”

11. Discussion Line 177. Which studies? Please include references

Author Reply:

Thanks for your advice. We have added the reference in the manuscript.

12. Discussion Lines 193-194. I’m not sure the urinary system has more protective barriers than the respiratory system. This sentence sounds strange. Would re-word.

Author Reply:

Thanks for your advice. We have revised the manuscript accordingly as follows.

“In contrast, those infections with multiple protective barriers (such as skin and musculoskeletal infections) had lower mortality rates (26).” (Section of Discussion, line 221-223)

13. Discussion Line 193. Remove period after mortality

Author Reply:

Thanks for your advice. We have corrected the manuscript accordingly.

14. Discussion Line 203. Suggest re-writing. What does tolerate mean?

Author Reply:

Thanks for your advice. We have re-written it.

“Those patient with lower risk of mortality might benefit from early de-escalation of the antibiotics.” (Section of Discussion, line 230-231)

15. Discussion Line 211. Change HER to HER

Author Reply:

Thanks for your advice. We have corrected the manuscript accordingly.

16. Conclusion Lines 220-225. This feels vague. What specific trend do you want readers to take away and what specific future advancement does this study help with?

Author Reply:

Thank you for your comment. We have revised the manuscript.

“There is a significant difference in the trend of incidence and outcome of sepsis from different anatomic sites of infection. Clinician should be aware of different anatomic sites of infection could cause higher mortality in septic patients such as intra-abdominal infection, lower respiratory tract infection, and biliary tract infection” (Section of Conclusion, line 259-262)

“Conclusions: The anatomic site of infection does have a differential impact on the mortality of septic patients. Intra-abdominal infection, lower respiratory tract infection, and biliary tract infection are associated with higher mortality in septic patients.” (Section of Abstract, line 49-51)

17. Tables 1, 2, and 3. Are these all weighted numbers? Please clarify. Suggest changing to title to reflect weighted analysis

Author Reply:

Thanks for your advice. These are weighted numbers. We have revised the table title accordingly.

18. Table 1. What comorbidity score was chosen? Please reference in methods.

Author Reply:

Thanks for your advice. We have added the scoring index into the manuscript accordingly.

“We use Elixhauser comorbidity Index as our comorbidity index.” (Section of Method, line 110-111)

19. Table 2. Change “systematic” to systemic. Double check throughout.

Author Reply:

Thanks for your advice. We have corrected the table accordingly.

20. Figure 1. Change “miss” to missing values

Author Reply:

Thanks for your advice. We have corrected the figure accordingly.

21. Figure 2b. Y axis title is not fully seen

Author Reply:

Thanks for your advice. We have corrected the figure accordingly.

22. Figures. Suggest placing legends underneath graphs

Author Reply:

Thanks for your advice. We have corrected the figure accordingly.

23. Figure 4. I like this.

Author Reply:

Thanks for your comment.

24. Supplementary Table 1. Why was primary bacteremia chosen as the reference? Primary bacteremia without a source seems like a rare entity and as such should not be the comparison.

Author Reply:

Thanks for your advice. Primary bacteremia was used as a reference because patient with only primary bacteremia had the lowest mortality. In order to make the chart more readable, we choose the lowest mortality one as the reference.

25. Supplementary Table 1. Almost all variables reach statistical significance due to the sample size. Is there a comparison population you could track over the time period (i.e. those requiring mechanical ventilation) to confidently say the sepsis trend is changing rather than an artifact of the large sample size? This could be a “control group analysis” throughout the methods, results, and discussion.

Author Reply:

Thank you for your comment. In the Supplementary Table 1, we focus on the mortality differences among different sites of infection. As for the trend of sepsis, there has been a debate in the trend of sepsis from prior literatures. (2-11) Rhee et al (12) found that the annul increase of sepsis was higher when using discharge code versus clinical criteria. So far Rhee et al (JAMA 2017) provided the best evidence that “neither the incidence of sepsis nor the combined outcome of death or discharge to hospice changed significantly.” (11) However, the primary aim of our study was to delineate the change in the incidence and in-hospital mortality between specific infection sites in sepsis over time, not the overall incidence of sepsis. Meanwhile, due to the limitation from our database, we are not able to provide such sophisticated results on this issue or control group analysis. We have listed it in the limitation section accordingly. Lastly, as we focused on the comparisons among different sites of infection, the results should not be affected significantly by the overall trend of sepsis or the definitive criteria applied. We have added this to our limitation as follows:

“Meanwhile, due to the insufficient information from the database, we cannot identify patients’ socioeconomical status, community-acquired or nosocomial, medical or surgical hospitalization. Also, we didn’t perform the control group analysis due to insufficient data. Further study involving more detailed in-patient data is required.” (Section of Discussion, line 246-249)

________________________________________

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: Yes: Matthew K. Hensley

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files to be viewed.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at figures@plos.org. Please note that Supporting Information files do not need this step.

1. Angus DC, Linde-Zwirble WT, Lidicker J, Clermont G, Carcillo J, Pinsky MR. Epidemiology of severe sepsis in the United States: analysis of incidence, outcome, and associated costs of care. Crit Care Med 2001; 29: 1303-1310.

2. Jolley RJ, Sawka KJ, Yergens DW, Quan H, Jette N, Doig CJ. Validity of administrative data in recording sepsis: a systematic review. Crit Care 2015; 19: 139.

3. Zahar JR, Timsit JF, Garrouste-Orgeas M, Francais A, Vesin A, Descorps-Declere A, Dubois Y, Souweine B, Haouache H, Goldgran-Toledano D, Allaouchiche B, Azoulay E, Adrie C. Outcomes in severe sepsis and patients with septic shock: pathogen species and infection sites are not associated with mortality. Crit Care Med 2011; 39: 1886-1895.

4. Stevenson EK, Rubenstein AR, Radin GT, Wiener RS, Walkey AJ. Two decades of mortality trends among patients with severe sepsis: a comparative meta-analysis*. Crit Care Med 2014; 42: 625-631.

5. Iwashyna TJ, Odden A, Rohde J, Bonham C, Kuhn L, Malani P, Chen L, Flanders S. Identifying patients with severe sepsis using administrative claims: patient-level validation of the angus implementation of the international consensus conference definition of severe sepsis. Med Care 2014; 52: e39-43.

6. Dombrovskiy VY, Martin AA, Sunderram J, Paz HL. Rapid increase in hospitalization and mortality rates for severe sepsis in the United States: a trend analysis from 1993 to 2003. Crit Care Med 2007; 35: 1244-1250.

7. Gaieski DF, Edwards JM, Kallan MJ, Carr BG. Benchmarking the incidence and mortality of severe sepsis in the United States. Crit Care Med 2013; 41: 1167-1174.

8. Lagu T, Rothberg MB, Shieh MS, Pekow PS, Steingrub JS, Lindenauer PK. Hospitalizations, costs, and outcomes of severe sepsis in the United States 2003 to 2007. Crit Care Med 2012; 40: 754-761.

9. Kumar G, Kumar N, Taneja A, Kaleekal T, Tarima S, McGinley E, Jimenez E, Mohan A, Khan RA, Whittle J, Jacobs E, Nanchal R. Nationwide trends of severe sepsis in the 21st century (2000-2007). Chest 2011; 140: 1223-1231.

10. Martin GS, Mannino DM, Eaton S, Moss M. The epidemiology of sepsis in the United States from 1979 through 2000. N Engl J Med 2003; 348: 1546-1554.

11. Rhee C, Dantes R, Epstein L, Murphy DJ, Seymour CW, Iwashyna TJ, Kadri SS, Angus DC, Danner RL, Fiore AE, Jernigan JA, Martin GS, Septimus E, Warren DK, Karcz A, Chan C, Menchaca JT, Wang R, Gruber S, Klompas M. Incidence and Trends of Sepsis in US Hospitals Using Clinical vs Claims Data, 2009-2014. Jama 2017; 318: 1241-1249.

12. Rhee C, Murphy MV, Li L, Platt R, Klompas M. Improving documentation and coding for acute organ dysfunction biases estimates of changing sepsis severity and burden: a retrospective study. Crit Care 2015; 19: 338.

18 Nov 2019

PONE-D-19-15923R1

Incidence, Trends, and Outcomes of Infection Sites among Hospitalizations of Sepsis: a Nationwide Study

PLOS ONE

Dear Dr Lee,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Please make sure to address the additional comments raised by the reviewers. In particular, please justify the use of odds / odds ratios for risk prediction  given its known limitations (e.g., Pepe MS, Am J Epidemiol 2004). Furthermore, please include sensitivity analyses using alternative sepsis coding strategy (e.g., 'Angus methodology') in the supplement section.

We would appreciate receiving your revised manuscript by Jan 02 2020 11:59PM. When you are ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter.

To enhance the reproducibility of your results, we recommend that if applicable you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). This letter should be uploaded as separate file and labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. This file should be uploaded as separate file and labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. This file should be uploaded as separate file and labeled 'Manuscript'.

Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.

We look forward to receiving your revised manuscript.

Kind regards,

Florian B. Mayr

Academic Editor

PLOS ONE

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: Chou et al. have significantly improved the methodology of this study. The interpretation of results is sound with expanded limitations and discussion. I agree with publication of these important results with minor revisions listed below.

Minor revisions:

1) Line 151-152: This sentence is confusing. Please re-word. “Male patients tend to be more likely to sepsis in all subperiods.”

2) Lines 182-189: You use Odds Ratios, but in the methods section lines 133-134 you mention that odds ratios are biased measures given the prevalence of sepsis and therefore do not reliably predict risk. Please clarify.

3) Lines 209-210: “One example being the surviving sepsis campaign being started at the beginning of this study who could account for increased incidence of sepsis”. This is confusing. Please re-word.

4) Lines 214-215: “which is common etiologic agent of pneumonia.” Please fix grammar.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: Yes: Matthew K Hensley

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files to be viewed.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at figures@plos.org. Please note that Supporting Information files do not need this step.

Submitted filename: Chou et al.docx

Click here for additional data file.

2 Dec 2019

Ref: PONE-D-19-15923

Title: Incidence, trends, and outcomes of infection sites among hospitalizations of sepsis: a nationwide study

Journal: PLOS ONE

Dear editor and reviewers:

Thank you for your e-mail with the referees’ comments. We have revised the manuscript as advised. The detailed point-by-point responses are as follows. We hope the manuscript is now acceptable to PLOS ONE.

Editor’s Comment:

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Please make sure to address the additional comments raised by the reviewers. In particular, please justify the use of odds / odds ratios for risk prediction given its known limitations (e.g., Pepe MS, Am J Epidemiol 2004). Furthermore, please include sensitivity analyses using alternative sepsis coding strategy (e.g., 'Angus methodology') in the supplement section.

Author Reply:

Thanks for your comment. To evaluate the impact of individual site of infection on the survival of sepsis patients, we fit a multivariable logistic regression model adjusting for age, sex, and comorbidity measures. We used the entire study period for this regression analysis to ensure adequate power to make reliable estimates of risk. Because the mortality rate for patients with sepsis is higher than 10% in this analysis, the rare disease assumption does not hold. As a result, risk ratios cannot be estimated by odds ratios. Therefore, we used the formula proposed by Zhang and Yu to approximate the relative risk. (1) We have corrected our manuscript and figures accordingly.

“As a result, risk ratios cannot be estimated by odds ratios.” (Line 134 in Section of Methods)

“Figure 4 shows the adjusted relative risk with 95% confidence intervals of infection site on the outcome of sepsis. Using primary bacteremia as reference, sepsis patients with intra-abdominal infection had the highest mortality (RR:4.21), followed by lower respiratory tract infection (RR: 3.84), biliary tract infection (RR: 3.24), systemic fungal infection (RR: 2.77), skin and skin structure infection (RR: 2.29), musculoskeletal infection (RR: 2.27), genitourinary traction infection (RR:2.19), or catheter related bloodstream infection (RR:2.15). Sensitivity analysis with Angus criteria showed similar trend of sepsis as our main results (Table S4-7 and Fig S1-3)

Fig 4. Survival impact of individual infection site in relation to primary bacteremia. The risk estimates were adjusted for all covariates listed in supporting table 1. RR refers to the relative risk. LCL and UCL refer to lower and upper confidence limits, respectively.” (Line 182-191 in Section of Results)

“Supporting Figure 3. Sensitivity Test - Survival impact of individual infection site in relation to primary bacteremia. The risk estimates were adjusted for all covariates listed in Supporting table 7. RR refers to the relative risk. LCL and UCL refer to lower and upper confidence limits, respectively.

Supporting Table 1. Covariates with associated relative risk in the outcome regression model.” (Line 371-376 in Section of Supporting Information)

“Supporting Table 7. Sensitivity Test - Covariates with associated relative risk in the outcome regression model.” (Line 386-388 in Section of Supporting Information)

We have included sensitivity analyses using Angus methodology in Table S4-7 and Fig S1-3.

We would appreciate receiving your revised manuscript by Jan 02 2020 11:59PM. When you are ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter.

To enhance the reproducibility of your results, we recommend that if applicable you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

Please include the following items when submitting your revised manuscript:

• A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). This letter should be uploaded as separate file and labeled 'Response to Reviewers'.

• A marked-up copy of your manuscript that highlights changes made to the original version. This file should be uploaded as separate file and labeled 'Revised Manuscript with Track Changes'.

• An unmarked version of your revised paper without tracked changes. This file should be uploaded as separate file and labeled 'Manuscript'.

Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.

We look forward to receiving your revised manuscript.

Kind regards,

Florian B. Mayr

Academic Editor

PLOS ONE

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: All comments have been addressed

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: Chou et al. have significantly improved the methodology of this study. The interpretation of results is sound with expanded limitations and discussion. I agree with publication of these important results with minor revisions listed below.

Minor revisions:

1) Line 151-152: This sentence is confusing. Please re-word. “Male patients tend to be more likely to sepsis in all subperiods.”

Author Reply:

Thanks for your comment. We have re-worded accordingly. “There are more male patients than female in all subperiods.” (Line 151-152 in Section of Results)

2) Lines 182-189: You use Odds Ratios, but in the methods section lines 133-134 you mention that odds ratios are biased measures given the prevalence of sepsis and therefore do not reliably predict risk. Please clarify.

Author Reply:

Thanks for your comment. To evaluate the impact of individual site of infection on the survival of sepsis patients, we fit a multivariable logistic regression model adjusting for age, sex, and comorbidity measures. We used the entire study period for this regression analysis to ensure adequate power to make reliable estimates of risk. Because the mortality rate for patients with sepsis is higher than 10% in this analysis, the rare disease assumption does not hold. As a result, risk ratios cannot be estimated by odds ratios. Therefore, we used the formula proposed by Zhang and Yu to approximate the relative risk. (1) We have corrected our manuscript, tables and figures accordingly.

“As a result, risk ratios cannot be estimated by odds ratios.” (Line 134 in Section of Methods)

“Figure 4 shows the adjusted relative risk with 95% confidence intervals of infection site on the outcome of sepsis. Using primary bacteremia as reference, sepsis patients with intra-abdominal infection had the highest mortality (RR:4.21), followed by lower respiratory tract infection (RR: 3.84), biliary tract infection (RR: 3.24), systemic fungal infection (RR: 2.77), skin and skin structure infection (RR: 2.29), musculoskeletal infection (RR: 2.27), genitourinary traction infection (RR:2.19), or catheter related bloodstream infection (RR:2.15). Sensitivity analysis with Angus criteria showed similar trend of sepsis as our main results (Table S4-7 and Fig S1-3)

Fig 4. Survival impact of individual infection site in relation to primary bacteremia. The risk estimates were adjusted for all covariates listed in supporting table 1. RR refers to the relative risk. LCL and UCL refer to lower and upper confidence limits, respectively.” (Line 182-191 in Section of Results)

“Supporting Figure 3. Sensitivity Test - Survival impact of individual infection site in relation to primary bacteremia. The risk estimates were adjusted for all covariates listed in Supporting table 7. RR refers to the relative risk. LCL and UCL refer to lower and upper confidence limits, respectively.

Supporting Table 1. Covariates with associated relative risk in the outcome regression model.” (Line 371-376 in Section of Supporting Information)

“Supporting Table 7. Sensitivity Test - Covariates with associated relative risk in the outcome regression model.” (Line 386-388 in Section of Supporting Information)

3) Lines 209-210: “One example being the surviving sepsis campaign being started at the beginning of this study who could account for increased incidence of sepsis”. This is confusing. Please re-word.

Author Reply:

Thanks for your comment. We have re-worded accordingly.

“For example, the surviving sepsis campaign started at the beginning of this study could account for early sepsis recognition and decreasing mortality with early antibiotic administration and three-hour bundle therapy (20-22).” (Line 208-210 in Section of Discussions)

4) Lines 214-215: “which is common etiologic agent of pneumonia.” Please fix grammar.

Author Reply:

Thanks for your comment. We have re-worded accordingly.

“Currently preventative strategies mostly aimed at streptococcus pneumoniae, which is the common pathogen of pneumonia.” (Line 212-213 in Section of Discussions)

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: Yes: Matthew K Hensley

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files to be viewed.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at figures@plos.org. Please note that Supporting Information files do not need this step.

Reference:

1. Zhang J, Yu KF. What's the relative risk? A method of correcting the odds ratio in cohort studies of common outcomes. Jama 1998; 280: 1690-1691.

Submitted filename: Response to Reviewers 2nd revision.doc

Click here for additional data file.

30 Dec 2019

Incidence, Trends, and Outcomes of Infection Sites among Hospitalizations of Sepsis: a Nationwide Study

PONE-D-19-15923R2

Dear Dr. Lee,

We are pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it complies with all outstanding technical requirements.

Within one week, you will receive an e-mail containing information on the amendments required prior to publication. When all required modifications have been addressed, you will receive a formal acceptance letter and your manuscript will proceed to our production department and be scheduled for publication.

Shortly after the formal acceptance letter is sent, an invoice for payment will follow. To ensure an efficient production and billing process, please log into Editorial Manager at https://www.editorialmanager.com/pone/, click the "Update My Information" link at the top of the page, and update your user information. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to enable them to help maximize its impact. If they will be preparing press materials for this manuscript, you must inform our press team as soon as possible and no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

With kind regards,

Florian B. Mayr

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

6 Jan 2020

PONE-D-19-15923R2

Incidence, Trends, and Outcomes of Infection Sites among Hospitalizations of Sepsis: a Nationwide Study

Dear Dr. Lee:

I am pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please notify them about your upcoming paper at this point, to enable them to help maximize its impact. If they will be preparing press materials for this manuscript, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

For any other questions or concerns, please email plosone@plos.org.

Thank you for submitting your work to PLOS ONE.

With kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Florian B. Mayr

Academic Editor

PLOS ONE