Jack Zhenhe Zhang
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Research Interests
Jack Zhenhe Zhang has a total of 41 co-authors. Here are the top co-authors by number of shared publications:
- Anna Lee
Department of Anaesthesia and Intensive Care, Chinese University of Hong Kong, Shatin, Hong Kong
4 shared publications Alwin Wai Tak Yeung
Department of Medicine and Geriatrics, Ruttonjee and Tang Shiu Kin Hospitals , Hong Kong SAR
3 shared publicationsJacky Ka Hing Chan
Intensive Care Unit, Tseung Kwan O Hospital , Hong Kong SAR
3 shared publications- Lowell Ling
Department of Anaesthesia and Intensive Care, The Chinese University of Hong Kong, Hong Kong SAR, China
3 shared publications Manimala Dharmangadan
Department of Intensive Care, Princess Margaret Hospital , Hong Kong SAR
3 shared publicationsTing Liong
Department of Intensive Care, United Christian Hospital , Hong Kong SAR
3 shared publications
Latest Publications
- Open Access
- 17 December 2025
- Published by Oxford University Press (OUP) in Open Forum Infectious Diseases
- Vol. 13 (1) , ofaf752
Regional Adherence to Early Sepsis Management Bundle and Associated Mortality in Hong Kong Between 2009–2018
Background: Longitudinal data are scarce on sepsis bundle adherence and associated survival at a country or regional level. Methods: A population-based electronic health record database was leveraged to determine temporal trends in sepsis bundle adherence (empirical broad-spectrum antibiotic administration, blood culture collection, lactate measurement) on sepsis onset day and antimicrobial resistance (AMR) prevalence. This study included all adult hospitalizations for community-acquired sepsis at 41 publicly funded hospitals in Hong Kong between 2009 and 2018. Generalized estimating equations were used to assess the association between full bundle adherence and its individual elements with hospital mortality. Results: Among 421 096 cases of community-acquired sepsis, the full bundle adherence rate increased from 0.2% in 2009 to 1.2% in 2018 (relative +18.9%/y, P < .001), with limited uptake of each element. The relative increase in empirical broad-spectrum antibiotics administration (+9.8%/y [95% CI, 8.3%–11.2%]) was faster than the AMR prevalence (+5.2%/y [95% CI, 3.6%–6.9%]). Full bundle adherence was associated with reduced mortality (adjusted odds ratio [ORadj], 0.75 [95% CI, .65–.86]). Blood culture collection was associated with reduced mortality (ORadj, 0.88 [95% CI, .83–.93]), while lactate measurement was associated only with reduced mortality in septic shock (ORadj, 0.85 [95% CI, .76–.94]). Broad-spectrum antibiotics was associated with reduced mortality (ORadj, 0.73 [95% CI, .56–.96]) when used appropriately in bacteremia from extended-spectrum beta-lactamase pathogens or methicillin-resistant Staphylococcus aureus. Conclusions: Basic sepsis care implementation remains challenging even in high-income settings. Empirical broad-spectrum antibiotic usage has outpaced AMR risk. Full sepsis bundle adherence was associated with improved survival, but empirical broad-spectrum antibiotics was associated with better survival only if used appropriately. Efforts should focus not only on ensuring bundle adherence but also on prioritizing the right treatments for the right patients.
- 17 December 2025
- Published by Oxford University Press (OUP) in Open Forum Infectious Diseases
- Vol. 13 (1) , ofaf752
Open AccessRegional Adherence to Early Sepsis Management Bundle and Associated Mortality in Hong Kong Between 2009–2018
Background: Longitudinal data are scarce on sepsis bundle adherence and associated survival at a country or regional level. Methods: A population-based electronic health record database was leveraged to determine temporal trends in sepsis bundle adherence (empirical broad-spectrum antibiotic administration, blood culture collection, lactate measurement) on sepsis onset day and antimicrobial resistance (AMR) prevalence. This study included all adult hospitalizations for community-acquired sepsis at 41 publicly funded hospitals in Hong Kong between 2009 and 2018. Generalized estimating equations were used to assess the association between full bundle adherence and its individual elements with hospital mortality. Results: Among 421 096 cases of community-acquired sepsis, the full bundle adherence rate increased from 0.2% in 2009 to 1.2% in 2018 (relative +18.9%/y, P < .001), with limited uptake of each element. The relative increase in empirical broad-spectrum antibiotics administration (+9.8%/y [95% CI, 8.3%–11.2%]) was faster than the AMR prevalence (+5.2%/y [95% CI, 3.6%–6.9%]). Full bundle adherence was associated with reduced mortality (adjusted odds ratio [ORadj], 0.75 [95% CI, .65–.86]). Blood culture collection was associated with reduced mortality (ORadj, 0.88 [95% CI, .83–.93]), while lactate measurement was associated only with reduced mortality in septic shock (ORadj, 0.85 [95% CI, .76–.94]). Broad-spectrum antibiotics was associated with reduced mortality (ORadj, 0.73 [95% CI, .56–.96]) when used appropriately in bacteremia from extended-spectrum beta-lactamase pathogens or methicillin-resistant Staphylococcus aureus. Conclusions: Basic sepsis care implementation remains challenging even in high-income settings. Empirical broad-spectrum antibiotic usage has outpaced AMR risk. Full sepsis bundle adherence was associated with improved survival, but empirical broad-spectrum antibiotics was associated with better survival only if used appropriately. Efforts should focus not only on ensuring bundle adherence but also on prioritizing the right treatments for the right patients.
- 12 September 2025
- Published by Wiley in Cochrane Database of Systematic Reviews
- Vol. 2025 (9) , CD003281
Stimulation of the wrist acupuncture point PC6 for preventing postoperative nausea and vomiting: a network meta-analysis
Postoperative nausea and vomiting (PONV) are common complications following surgery and anaesthesia. Antiemetic drugs are only partially effective in preventing PONV. Stimulating the PC6 acupoint(s) on the wrist offers an alternative approach, but the effectiveness of the various common techniques is unclear. To update and compare the effects and safety of PC6 acupoint stimulation with or without antiemetic drug(s) versus sham or antiemetic drug(s) for preventing postoperative nausea (PON) and postoperative vomiting (POV) in people undergoing surgery, and to identify the most effective techniques using network meta-analyses (NMAs). We searched CENTRAL, MEDLINE, Embase, ISI Web of Science, CINAHL, WHO Global Index Medicus, major trial registries, and reference lists of articles for studies up to 6 June 2025, with no language restrictions. Parallel randomised controlled trials of techniques that compared any combinations of PC6 acupoint stimulation, drug therapy, and sham for preventing PONV in children and adults were eligible. Interventions included invasive techniques (e.g. needle acupuncture) and noninvasive techniques (e.g. acupressure wristbands). Drug therapy included antiemetics that belonged to one of the following substance classes: 5-HT3 receptor antagonists, D2 receptor antagonists, corticosteroids, and antihistamines. The sham group included sham PC6 interventions and/or placebo saline antiemetic interventions. Critical outcomes: incidence of PON and POV. Important outcomes: need for rescue antiemetic therapy when prophylaxis failed and adverse events (side effects). We assessed the risk of bias in the included studies using RoB 1. We performed frequentist NMAs using random-effects models to report risk ratios (RRs) with 95% confidence intervals (95% CIs). We compared the relative effects (with sham as reference) of six broad classes of PC6 acupoint stimulation techniques and their combined use with antiemetic drugs on the outcomes. We summarised the safety data narratively due to heterogeneous reporting of adverse events (side effects). We assessed the certainty of evidence of the NMA treatment effect on outcomes according to the CINeMA and GRADE approaches. This update included 77 trials, conducted between 1986 and 2022, involving 9847 participants. The majority were adults across several countries (USA, South Korea, India, China, Turkey, and Iran). There were 58 (33.9%) sham groups, 16 (9.4%) invasive PC6 acupoint stimulation groups, 50 (29.3%) noninvasive PC6 acupoint stimulation groups, 32 (18.7%) antiemetic(s) groups, 4 (2.3%) combined invasive PC6 acupoint stimulation and antiemetic(s) groups, 10 (5.8%) combined noninvasive PC6 acupoint stimulation and antiemetic(s) groups, and one (0.6%) combined invasive and noninvasive PC6 acupoint stimulation group for the NMAs. The high risk of bias was primarily due to selective reporting. Postoperative nausea The evidence from the NMA (63 studies; 7534 participants) suggests that the combined invasive PC6 and antiemetics technique probably results in a larger reduction in PON than sham (RR 0.21, 95% CI 0.10 to 0.45; moderate confidence from indirect evidence only). The evidence also suggests that invasive PC6, noninvasive PC6, and antiemetic(s) may reduce PON, compared to sham (invasive PC6: RR 0.49, 95% CI 0.38 to 0.64, low confidence; noninvasive PC6: RR 0.67, 95% CI 0.58 to 0.76, low confidence; antiemetic(s): RR 0.73, 95% CI 0.59 to 0.91, very low confidence). The combined use of noninvasive PC6 and antiemetic(s) may reduce PON, compared to sham, based on indirect evidence (RR 0.65, 95% CI 0.45 to 0.96; low confidence). Postoperative vomiting The evidence from the NMA (75 studies; 8627 participants) suggests that the combined invasive PC6 and antiemetics technique may result in a moderate reduction in POV, compared to sham (RR 0.37, 95% CI 0.18 to 0.76; low confidence from indirect evidence only). The evidence also suggests that invasive PC6, noninvasive PC6, antiemetic(s), and combined noninvasive PC6 and antiemetic(s) may reduce POV, compared to sham (invasive PC6: RR 0.47, 95% CI 0.34 to 0.64, low confidence; noninvasive PC6: RR 0.58, 95% CI 0.49 to 0.70, low confidence; antiemetic(s): RR 0.62, 95% CI 0.47 to 0.81, very low confidence; combined noninvasive PC6 and antiemetic(s): RR 0.52, 95% CI 0.33 to 0.83, very low confidence). The combined use of noninvasive PC6 and invasive PC6 may result in little to no difference in POV, compared to sham, based on direct evidence (RR 1.02, 95% CI 0.43 to 2.44; low confidence). Need for rescue antiemetic drug(s) The evidence from the NMA (55 studies; 6614 participants) suggests that combined use of noninvasive PC6 and antiemetic(s) probably reduces the need for rescue antiemetic drug(s) when prophylactic techniques fail, compared to sham (RR 0.44, 95% CI 0.28 to 0.70; moderate confidence from indirect evidence). Antiemetic(s), noninvasive PC6, and invasive PC6 may reduce the need for rescue antiemetics, compared to sham (antiemetic(s): RR 0.56, 95% CI 0.40 to 0.80, very low confidence; noninvasive PC6: RR 0.60, 95% CI 0.51 to 0.70, low confidence; invasive PC6: RR 0.61, 95% CI 0.44 to 0.84, low confidence). The combined use of invasive PC6 and antiemetic(s) (RR 0.31, 95% CI 0.04 to 2.49; low confidence from indirect evidence only) and the combined use of noninvasive PC6 and invasive PC6 may result in little or no difference in the need for rescue antiemetic drug(s), compared to sham (RR 1.41, 95% CI 0.64 to 3.09; low confidence from direct evidence only). Adverse events (side effects) None of the included studies reported serious or long-term complications. Thirty-nine studies, involving 5334 participants, reported minor side effects that go away (e.g. skin irritation, redness, and swelling) with PC6 acupoint stimulation, but the evidence is very uncertain. Publication bias was not apparent in the NMA funnel plots. NMAs suggest that both invasive and noninvasive PC6 acupoint stimulation combined with antiemetics may reduce PON and POV, compared to sham, and likely lower rescue antiemetic use. PC6 acupoint stimulation may cause minor side effects, but the evidence is very uncertain. This updated Cochrane review was partially funded by the Cochrane Complementary Medicine Field Bursary 2020 from the National Center for Complementary and Alternative Medicine, National Institutes of Health (Grant Number R24 AT001293). Protocol (2001): doi.org/10.1002/14651858.CD003281 Original review (2004): doi.org/10.1002/14651858.CD003281.pub2 Update review (2009): doi.org/10.1002/14651858.CD003281.pub3 Update review (2015): doi.org/10.1002/14651858.CD003281.pub4.
- 12 September 2025
- Published by Wiley in Cochrane Database of Systematic Reviews
- Vol. 2025 (9) , CD003281
Stimulation of the wrist acupuncture point PC6 for preventing postoperative nausea and vomiting: a network meta-analysis
Postoperative nausea and vomiting (PONV) are common complications following surgery and anaesthesia. Antiemetic drugs are only partially effective in preventing PONV. Stimulating the PC6 acupoint(s) on the wrist offers an alternative approach, but the effectiveness of the various common techniques is unclear. To update and compare the effects and safety of PC6 acupoint stimulation with or without antiemetic drug(s) versus sham or antiemetic drug(s) for preventing postoperative nausea (PON) and postoperative vomiting (POV) in people undergoing surgery, and to identify the most effective techniques using network meta-analyses (NMAs). We searched CENTRAL, MEDLINE, Embase, ISI Web of Science, CINAHL, WHO Global Index Medicus, major trial registries, and reference lists of articles for studies up to 6 June 2025, with no language restrictions. Parallel randomised controlled trials of techniques that compared any combinations of PC6 acupoint stimulation, drug therapy, and sham for preventing PONV in children and adults were eligible. Interventions included invasive techniques (e.g. needle acupuncture) and noninvasive techniques (e.g. acupressure wristbands). Drug therapy included antiemetics that belonged to one of the following substance classes: 5-HT3 receptor antagonists, D2 receptor antagonists, corticosteroids, and antihistamines. The sham group included sham PC6 interventions and/or placebo saline antiemetic interventions. Critical outcomes: incidence of PON and POV. Important outcomes: need for rescue antiemetic therapy when prophylaxis failed and adverse events (side effects). We assessed the risk of bias in the included studies using RoB 1. We performed frequentist NMAs using random-effects models to report risk ratios (RRs) with 95% confidence intervals (95% CIs). We compared the relative effects (with sham as reference) of six broad classes of PC6 acupoint stimulation techniques and their combined use with antiemetic drugs on the outcomes. We summarised the safety data narratively due to heterogeneous reporting of adverse events (side effects). We assessed the certainty of evidence of the NMA treatment effect on outcomes according to the CINeMA and GRADE approaches. This update included 77 trials, conducted between 1986 and 2022, involving 9847 participants. The majority were adults across several countries (USA, South Korea, India, China, Turkey, and Iran). There were 58 (33.9%) sham groups, 16 (9.4%) invasive PC6 acupoint stimulation groups, 50 (29.3%) noninvasive PC6 acupoint stimulation groups, 32 (18.7%) antiemetic(s) groups, 4 (2.3%) combined invasive PC6 acupoint stimulation and antiemetic(s) groups, 10 (5.8%) combined noninvasive PC6 acupoint stimulation and antiemetic(s) groups, and one (0.6%) combined invasive and noninvasive PC6 acupoint stimulation group for the NMAs. The high risk of bias was primarily due to selective reporting. Postoperative nausea The evidence from the NMA (63 studies; 7534 participants) suggests that the combined invasive PC6 and antiemetics technique probably results in a larger reduction in PON than sham (RR 0.21, 95% CI 0.10 to 0.45; moderate confidence from indirect evidence only). The evidence also suggests that invasive PC6, noninvasive PC6, and antiemetic(s) may reduce PON, compared to sham (invasive PC6: RR 0.49, 95% CI 0.38 to 0.64, low confidence; noninvasive PC6: RR 0.67, 95% CI 0.58 to 0.76, low confidence; antiemetic(s): RR 0.73, 95% CI 0.59 to 0.91, very low confidence). The combined use of noninvasive PC6 and antiemetic(s) may reduce PON, compared to sham, based on indirect evidence (RR 0.65, 95% CI 0.45 to 0.96; low confidence). Postoperative vomiting The evidence from the NMA (75 studies; 8627 participants) suggests that the combined invasive PC6 and antiemetics technique may result in a moderate reduction in POV, compared to sham (RR 0.37, 95% CI 0.18 to 0.76; low confidence from indirect evidence only). The evidence also suggests that invasive PC6, noninvasive PC6, antiemetic(s), and combined noninvasive PC6 and antiemetic(s) may reduce POV, compared to sham (invasive PC6: RR 0.47, 95% CI 0.34 to 0.64, low confidence; noninvasive PC6: RR 0.58, 95% CI 0.49 to 0.70, low confidence; antiemetic(s): RR 0.62, 95% CI 0.47 to 0.81, very low confidence; combined noninvasive PC6 and antiemetic(s): RR 0.52, 95% CI 0.33 to 0.83, very low confidence). The combined use of noninvasive PC6 and invasive PC6 may result in little to no difference in POV, compared to sham, based on direct evidence (RR 1.02, 95% CI 0.43 to 2.44; low confidence). Need for rescue antiemetic drug(s) The evidence from the NMA (55 studies; 6614 participants) suggests that combined use of noninvasive PC6 and antiemetic(s) probably reduces the need for rescue antiemetic drug(s) when prophylactic techniques fail, compared to sham (RR 0.44, 95% CI 0.28 to 0.70; moderate confidence from indirect evidence). Antiemetic(s), noninvasive PC6, and invasive PC6 may reduce the need for rescue antiemetics, compared to sham (antiemetic(s): RR 0.56, 95% CI 0.40 to 0.80, very low confidence; noninvasive PC6: RR 0.60, 95% CI 0.51 to 0.70, low confidence; invasive PC6: RR 0.61, 95% CI 0.44 to 0.84, low confidence). The combined use of invasive PC6 and antiemetic(s) (RR 0.31, 95% CI 0.04 to 2.49; low confidence from indirect evidence only) and the combined use of noninvasive PC6 and invasive PC6 may result in little or no difference in the need for rescue antiemetic drug(s), compared to sham (RR 1.41, 95% CI 0.64 to 3.09; low confidence from direct evidence only). Adverse events (side effects) None of the included studies reported serious or long-term complications. Thirty-nine studies, involving 5334 participants, reported minor side effects that go away (e.g. skin irritation, redness, and swelling) with PC6 acupoint stimulation, but the evidence is very uncertain. Publication bias was not apparent in the NMA funnel plots. NMAs suggest that both invasive and noninvasive PC6 acupoint stimulation combined with antiemetics may reduce PON and POV, compared to sham, and likely lower rescue antiemetic use. PC6 acupoint stimulation may cause minor side effects, but the evidence is very uncertain. This updated Cochrane review was partially funded by the Cochrane Complementary Medicine Field Bursary 2020 from the National Center for Complementary and Alternative Medicine, National Institutes of Health (Grant Number R24 AT001293). Protocol (2001): doi.org/10.1002/14651858.CD003281 Original review (2004): doi.org/10.1002/14651858.CD003281.pub2 Update review (2009): doi.org/10.1002/14651858.CD003281.pub3 Update review (2015): doi.org/10.1002/14651858.CD003281.pub4.
- Open Access
- 23 June 2024
- Published by Springer Nature in BMC Infectious Diseases
- Vol. 24 (1) , 1-12
Association between time to surgery and hospital mortality in patients with community-acquired limb necrotizing fasciitis: an 11-year multicenter retrospective cohort analysis
Background: Necrotizing fasciitis (NF) is a rare but potentially life-threatening soft tissue infection. The objective of this study was to assess the association between timely surgery within 6 h and hospital mortality in patients with limb NF, and to describe the trends in patients with NF, time to surgery and standardized mortality ratio (SMR) over 11 years. Methods: This was a multicenter, retrospective cohort study of all intensive care unit patients who had emergency surgery within 24 h of hospitalization for limb NF between April 1, 2008 and March 31, 2019 in Hong Kong. Timely surgery was defined as the first surgical treatment within 6 h of initial hospitalization. Appropriate antibiotics were achieved if the patient was given antibiotic(s) for all documented pathogens prior to or on day of culture results. The primary outcome was hospital mortality. Results: There were 495 patients (median age 62 years, 349 (70.5%) males) with limb NF treated by surgery within 24 h of hospitalization over the 11 years. Appropriate antibiotic(s) were used in 392 (79.2%) patients. There were 181 (36.5%) deaths. Timely surgery was not associated with hospital mortality (Relative Risk 0.89, 95% CI: 0.73 to 1.07) but admission year, advanced age, higher severity of illness, comorbidities, renal replacement therapy, vasopressor use, and type of surgery were significant predictors in the multivariable model. There was an upward trend in NF diagnosis (1.9 cases/year, 95% CI: 0.7 to 3.1; P < 0.01; R2 = 0.60) but there was no downward trend in median time to surgery (-0.2 h/year, 95% CI: -0.4 to 0.1; P = 0.16) or SMR (-0.02/year, 95% CI: -0.06 to 0.01; P = 0.22; R2 = 0.16). Conclusions: Among patients operated within 24 h, very early surgery within 6–12 h was not associated with survival. Increasing limb NF cases were reported each year but mortality remained high despite a high rate of appropriate antibiotic use and timely surgical intervention.
- 23 June 2024
- Published by Springer Nature in BMC Infectious Diseases
- Vol. 24 (1) , 1-12
Open AccessAssociation between time to surgery and hospital mortality in patients with community-acquired limb necrotizing fasciitis: an 11-year multicenter retrospective cohort analysis
Background: Necrotizing fasciitis (NF) is a rare but potentially life-threatening soft tissue infection. The objective of this study was to assess the association between timely surgery within 6 h and hospital mortality in patients with limb NF, and to describe the trends in patients with NF, time to surgery and standardized mortality ratio (SMR) over 11 years. Methods: This was a multicenter, retrospective cohort study of all intensive care unit patients who had emergency surgery within 24 h of hospitalization for limb NF between April 1, 2008 and March 31, 2019 in Hong Kong. Timely surgery was defined as the first surgical treatment within 6 h of initial hospitalization. Appropriate antibiotics were achieved if the patient was given antibiotic(s) for all documented pathogens prior to or on day of culture results. The primary outcome was hospital mortality. Results: There were 495 patients (median age 62 years, 349 (70.5%) males) with limb NF treated by surgery within 24 h of hospitalization over the 11 years. Appropriate antibiotic(s) were used in 392 (79.2%) patients. There were 181 (36.5%) deaths. Timely surgery was not associated with hospital mortality (Relative Risk 0.89, 95% CI: 0.73 to 1.07) but admission year, advanced age, higher severity of illness, comorbidities, renal replacement therapy, vasopressor use, and type of surgery were significant predictors in the multivariable model. There was an upward trend in NF diagnosis (1.9 cases/year, 95% CI: 0.7 to 3.1; P < 0.01; R2 = 0.60) but there was no downward trend in median time to surgery (-0.2 h/year, 95% CI: -0.4 to 0.1; P = 0.16) or SMR (-0.02/year, 95% CI: -0.06 to 0.01; P = 0.22; R2 = 0.16). Conclusions: Among patients operated within 24 h, very early surgery within 6–12 h was not associated with survival. Increasing limb NF cases were reported each year but mortality remained high despite a high rate of appropriate antibiotic use and timely surgical intervention.
- Open Access
- 19 August 2023
- Published by Oxford University Press (OUP) in Clinical Infectious Diseases
- Vol. 80 (1) , 91-100
Population Sepsis Incidence, Mortality, and Trends in Hong Kong Between 2009 and 2018 Using Clinical and Administrative Data
Sepsis surveillance using electronic health record (EHR)-based data may provide more accurate epidemiologic estimates than administrative data, but experience with this approach to estimate population-level sepsis burden is lacking. This was a retrospective cohort study including all adults admitted to publicly-funded hospitals in Hong Kong between 2009-2018. Sepsis was defined as clinical evidence of presumed infection (clinical cultures and treatment with antibiotics) and concurrent acute organ dysfunction (≥2 point increase in baseline SOFA score). Trends in incidence, mortality, and case fatality risk (CFR) were modelled by exponential regression. Performance of the EHR-based definition was compared with 4 administrative definitions using 500 medical record reviews. Among 13,550,168 hospital episodes during the study period, 485,057 (3.6%) had sepsis by EHR-based criteria with 21.5% CFR. In 2018, age- and sex-adjusted standardized sepsis incidence was 759 per 100,000 (relative +2.9%/year [95%CI 2.0, 3.8%] between 2009-2018) and standardized sepsis mortality was 156 per 100,000 (relative +1.9%/year [95%CI 0.9,2.9%]). Despite decreasing CFR (relative -0.5%/year [95%CI -1.0, -0.1%]), sepsis accounted for an increasing proportion of all deaths (relative +3.9%/year [95%CI 2.9, 4.9%]). Medical record reviews demonstrated that the EHR-based definition more accurately identified sepsis than administrative definitions (AUC 0.91 vs 0.52–0.55, p < 0.001). An objective EHR-based surveillance definition demonstrated an increase in population-level standardized sepsis incidence and mortality in Hong Kong between 2009-2018 and was much more accurate than administrative definitions. These findings demonstrate the feasibility and advantages of an EHR-based approach for widescale sepsis surveillance.
- 19 August 2023
- Published by Oxford University Press (OUP) in Clinical Infectious Diseases
- Vol. 80 (1) , 91-100
Open AccessPopulation Sepsis Incidence, Mortality, and Trends in Hong Kong Between 2009 and 2018 Using Clinical and Administrative Data
Sepsis surveillance using electronic health record (EHR)-based data may provide more accurate epidemiologic estimates than administrative data, but experience with this approach to estimate population-level sepsis burden is lacking. This was a retrospective cohort study including all adults admitted to publicly-funded hospitals in Hong Kong between 2009-2018. Sepsis was defined as clinical evidence of presumed infection (clinical cultures and treatment with antibiotics) and concurrent acute organ dysfunction (≥2 point increase in baseline SOFA score). Trends in incidence, mortality, and case fatality risk (CFR) were modelled by exponential regression. Performance of the EHR-based definition was compared with 4 administrative definitions using 500 medical record reviews. Among 13,550,168 hospital episodes during the study period, 485,057 (3.6%) had sepsis by EHR-based criteria with 21.5% CFR. In 2018, age- and sex-adjusted standardized sepsis incidence was 759 per 100,000 (relative +2.9%/year [95%CI 2.0, 3.8%] between 2009-2018) and standardized sepsis mortality was 156 per 100,000 (relative +1.9%/year [95%CI 0.9,2.9%]). Despite decreasing CFR (relative -0.5%/year [95%CI -1.0, -0.1%]), sepsis accounted for an increasing proportion of all deaths (relative +3.9%/year [95%CI 2.9, 4.9%]). Medical record reviews demonstrated that the EHR-based definition more accurately identified sepsis than administrative definitions (AUC 0.91 vs 0.52–0.55, p < 0.001). An objective EHR-based surveillance definition demonstrated an increase in population-level standardized sepsis incidence and mortality in Hong Kong between 2009-2018 and was much more accurate than administrative definitions. These findings demonstrate the feasibility and advantages of an EHR-based approach for widescale sepsis surveillance.