Vibrio infections in Aotearoa New Zealand (1998–2024) and the effect of current disease reporting structures

Vibrio spp. are motile, curved Gram-negative bacteria that naturally live in aquatic environments. Four species are important human pathogens because they frequently cause infections and/or severe health outcomes: V. cholerae, V. parahaemolyticus, V. vulnificus and V. alginolyticus.¹ Humans can be infected when they handle or consume shellfish or other seafood that have accumulated Vibrio spp., or come into contact with water containing these bacteria. Vibrio infections commonly present as gastrointestinal or soft tissue infections, but a range of factors, including host susceptibility and the infecting Vibrio spp., influence the clinical manifestation.¹

Toxigenic V. cholerae (serotypes O1 and O139), which usually inhabit freshwater and cause cholera,¹ are not endemic to Aotearoa New Zealand.² Domestically acquired infections of non-toxigenic V. cholerae have been reported in New Zealand.^3,4 V. parahaemolyticus is most often associated with foodborne gastroenteritis but can also infect soft tissues.¹ Internationally important V. parahaemolyticus subtypes have caused illness in New Zealand.⁵ Gastrointestinal infections from V. vulnificus can be self-limiting but can also rapidly progress to septicaemia in susceptible individuals, and half of these cases may die despite aggressive medical treatment.¹ V. vulnificus wound infections may require rapid debridement of necrotic tissue and/or amputation, which may still be insufficient to prevent sepsis and mortality.⁶ V. alginolyticus is commonly isolated from ear and wound infections.⁷

Seasonally, vibrio infections, or “vibriosis” (used here to describe non-cholera clinical manifestations) tend to peak during summer and early autumn months when the concentrations of Vibrio spp. in the environment increase in response to warmer waters.⁸ Longer temporal trends present a more troubling picture of climate change–driven sea temperature rise, expansion of pathogenic Vibrio spp. into global regions previously considered unsuitable for these bacteria and a concurrent increase in vibriosis incidence.^6,8–10 The sea surface temperatures along most of New Zealand’s coastal regions are increasing at rates of ≥0.23 degrees Celsius per decade, and since 2010 there has been a significant increase in the frequency, duration and intensity of marine heatwaves.¹¹ New Zealand studies have shown a positive correlation between the concentrations of V. parahaemolyticus and V. vulnificus in aquaculture shellfish and sea surface temperatures.^12,13 Since 2019, vibriosis outbreaks have been linked to seafood harvested from New Zealand’s coastal waters.¹⁴

Infections from Vibrio spp. are only notifiable in New Zealand if an infected person has: cholera, acute gastroenteritis (AG) caused by V. parahaemolyticus, an occupation that puts others at risk of AG or is part of an AG outbreak.¹⁵ There is no national reporting system for vibriosis that does not fit these parameters. Here, we describe vibriosis in New Zealand as revealed through collating available data, and show how de-centralised reporting constrains a cohesive public health response.

Methods

Data related to human infections of Vibrio spp. were collated from the following New Zealand sources:

• The national notifiable disease surveillance database, EpiSurv. Notified sporadic cases for the period 1998–2022 were provided in August 2023 by the New Zealand Institute for Public Health and Forensic Science (PHF Science).
• PHF Science’s Enteric Reference Laboratory (ERL). Clinical isolates of Vibrio spp. confirmed by the ERL during the period 2001–2022, from single-patient events, as of May 2023.
• National datasets of hospital discharges and mortalities. Health New Zealand – Te Whatu Ora provided extracts for the periods 2000–2024 (discharges) and 2000–2020 (mortalities) in February 2025 for the following International Statistical Classification of Diseases and Related Health Problems (ICD-10-AM/ACHI/ACS) codes: A00.0 (cholera due to V. cholerae O1, biovar cholerae/classical), A00.1 (cholera due to V. cholerae O1, biovar El Tor), A00.9 (cholera unspecified), A05.3 (foodborne V. parahaemolyticus intoxication) and B96.82 (V. vulnificus as the cause of diseases classified to other chapters).
• One regional diagnostic laboratory (RDL) that receives hospital and community clinical samples from a large North Island region that includes inland and coastal population centres. This laboratory collated internal data on samples that tested positive for Vibrio spp. during the period January 2008–December 2023, undertook their own analyses then contributed the numbers and supporting information required for this manuscript in April 2024.
• Publicly available reports describing V. parahaemolyticus outbreaks during the period 1998–2023.

Any changes made within these databases since provision of the extracts are not reflected in this report.

Ethics approval was not required from New Zealand’s Health and Disability Ethics Committee because identifiable data were not requested, and no attempt was made to link individual records between datasets. Access/use conditions set by the data owners/custodians were adhered to.

Collated data were examined for trends and data gaps, and how these were influenced by the data collection systems and data sources. Analysed data included the sex and prioritised ethnicity of each individual, which may or may not be self-reported and may or may not reflect an individual’s gender or ethnic identity preference. The ethnicity and regional distributions of the New Zealand resident population were sourced from the 2023 Census as reported by Stats NZ Tatauranga Aotearoa.¹⁶ The prioritised dates used to examine temporal trends were: EpiSurv, symptom onset date (75% records) then report date; ERL, sample collection date (96% records) then isolate receipt date; hospital discharges, event start date (100% records); RDL, sample receipt date (100% records).

Results

Data overview

Figure 1 illustrates how the data collated for this study were generated. Each database holds data at a national level except for diagnostic laboratories. The datasets overlap and, potentially, one hospitalised patient’s infection could be captured in all four databases. Resolving the overlap between these datasets would require access to personal data held by multiple agencies. The retrieved data were sufficient to indicate trends, and a decision was taken to proceed without resolving this overlap.

View Figure 1, Table 1–2.

Table 1 summarises key results. These data demonstrate that people of all ages can be affected by vibriosis. They also suggest that Pacific peoples and Māori are disproportionately affected by vibriosis compared with those identifying as European (c.f. New Zealand 2023 resident population 18% Māori, 9% Pacific peoples and 68% European), noting that ethnicity data are reported differently in each database.

Table 2 summarises 12 V. parahaemolyticus outbreaks reported in New Zealand between 1998 and 2022. Prior to 2019, there were no reported outbreaks linked to seafood harvested from New Zealand. Two outbreaks (2019, 2020) occurred during the early winter months.

The range of Vibrio spp. causing infections

Data provided by the ERL and RDL show that V. parahaemolyticus, V. alginolyticus and non-O1/non-O139 V. cholerae were the species most often identified from clinical samples, including from cases who had not travelled outside New Zealand during the incubation period. Other Vibrio spp. were also isolated from people who did not report travel, including V. mimicus, V. fluvialis, V. vulnificus, V. harveyi and V. hollisae. Toxigenic V. cholerae were infrequently confirmed, and only from people who reported overseas travel. The only relevant records obtained from the mortalities database were for cholera (unspecified) contributing to three deaths.

Clinical trends

The RDL data indicate the range of adverse health effects caused by Vibrio spp. Of the 326 vibrio-positive samples, the majority were indicative of ear infections (60% of samples were ear swabs), followed by gastroenteritis (20% were faecal samples) and other soft tissue infections (18% were wound, tissue or eye samples). The remaining 2% were blood cultures, suggestive of septicaemia.

The hospitalisation data highlight the seriousness of some infections. Of the 117 sporadic V. parahaemolyticus AG cases with known hospitalisation status reported to EpiSurv, 27% were hospitalised (no deaths were reported). The highest percentage hospitalised from the 12 V. parahaemolyticus outbreaks (Table 2) was 44% (seven out of 16 cases, 2020 outbreak). By hospital discharge data, V. vulnificus and V. cholerae infections also contributed to, or resulted in, 57 individuals being hospitalised. Of the 94 hospital discharge records, 35% included at least one clinical diagnosis code indicating concurrent diabetes mellitus as an underlying chronic health condition, with eight of these patients also having chronic kidney disease. Chronic viral hepatitis B without delta-agent was reported for a further 5% of records (three out of five included a V. vulnificus diagnosis). Acute kidney failure was a common presentation (reported in 37% of the records with and without codes indicative of diabetes).

V. parahaemolyticus

The RDL data indicate that this pathogen has a role in gastrointestinal, ear and soft tissue infections, since V. parahaemolyticus was confirmed in faecal (61%), ear (20%), wound (16%) and blood (3%) samples (n=75). Age data from EpiSurv, the RDL and the hospital discharges produced median ages of 47, 55 and 56 years, respectively, showing higher prevalence among middle-aged adults compared with other age groups. Of the 39 hospital discharge events associated with V. parahaemolyticus, this condition was the primary diagnosis in 69% of these events. Seven cases had co-infections with non-vibrio bacteria. Of these, three had wounds or lower limb cellulitis and, other than the “foodborne’” code for V. parahaemolyticus, no clinical coding suggestive of gastroenteritis. This suggests V. parahaemolyticus was part of a mixed culture obtained from a wound sample.

V. alginolyticus

RDL data show V. alginolyticus infections (n=156) were common among children and younger adults (72% were aged ≤40 years) and males (72% of patient episodes), and that the majority (78%) of the samples yielding V. alginolyticus were ear swabs (a further 21% were wound swabs). Other bacterial species were also identified from 54% of the ear swabs and 31% of the wound swabs (commonly Staphylococcus aureus); thus V. alginolyticus may not always be the primary cause of ear or skin infections. The ERL confirmed V. alginolyticus in 57 samples from people who did not report overseas travel, implying domestically acquired infections.

V. vulnificus

The hospital discharge dataset was most informative for V. vulnificus infections, noting that the relevant ICD-10 code (B96.82) was introduced in January 2014 and is typically used to support the primary diagnosis. Of the 23 discharge records associated with B96.82, 78% were male. The data indicate that 13 of the hospitalised patients had wounds, ulcerations and/or cellulitis, with some progressing to sepsis (seven patients) and/or necrotising fasciitis (three patients). Another six records suggest V. vulnificus was a cause of gastrointestinal infections rather than wound infections. The ERL confirmed V. vulnificus isolates from blood (two samples), faeces (one sample) and a leg ulcer sample (all four were from male patients who did not report overseas travel).

Non-O1/non-O139 V. cholerae

The ERL dataset was most informative for providing insights into domestically acquired infections, since 107/160 of the non-O1/non-O139 V. cholerae isolates were from people who had not reported travelling overseas during the incubation period. These 107 implied domestic cases were across a broad age range (5–86 years, median 52 years), and of the 105 cases with sex reported 61% were male. The majority (79%) of isolates from domestic cases were from faecal samples, suggesting gastrointestinal infection was an important clinical presentation. Isolates were also from ears (12%, 10/13 were male patients), blood (7%) and wounds (2%).

Temporal and geographical trends

The way infectious disease data are recorded in New Zealand did not allow for consistent adjustments that would account for the delay between patient exposure to Vibrio spp. or the appearance of symptomatic illness and the initiation of an entry into a database (see Methods). In the EpiSurv and ERL data, a patient’s symptom onset date or sample collection date was compared to the report month, and these results indicated that most reporting occurred within 1 month of symptomatic illness or sample collection (data not shown).

Based on the best dates available in each dataset (see Methods), infections have occurred throughout the year but there were more cases during January–March compared to other quarters. When excluding cases who had reported travelling overseas, 43% of the V. parahaemolyticus AG cases reported to EpiSurv became symptomatic or were notified during January–March (n=106), and 50% of the vibrio-positive samples were received at the ERL during these months (n=404). Of all RDL samples yielding Vibrio spp. (n=326, Table 1), 61% were received during January–March. Of the 94 hospital discharges, 45% had admission dates within this period.

Each dataset was examined for temporal trends in the annual number of records over time but no consistent trend was evident (data not shown).

EpiSurv location metadata shows that the highest number of V. parahaemolyticus cases were in the northern New Zealand regions of Auckland (17%) and Counties Manukau (18%) (c.f. New Zealand 2023 resident population 9% Auckland, 12% Counties Manukau). However, the location where each case was potentially exposed to Vibrio spp. may be different. The seafood collection sites reported by cases from the 2021–2022 V. parahaemolyticus outbreak (Table 2) were distributed along the length of New Zealand.²¹

Sources of infection

Food was the implicated source of infection for 71% (89/126) of the V. parahaemolyticus AG cases reported in EpiSurv. Cases reported consuming a wide range of aquatic foods but most commonly mussels and oysters, particularly in their raw state. Unless an outbreak is declared, or a specific food complaint is made to a relevant authority, there is no attempt to test food items for potential pathogens. There were 11 cases whose information suggested they had not travelled but had consumed seafood privately imported from a Pacific island, including a cluster of nine cases that occurred during the years 1999 and 2000. Outbreaks associated with privately imported seafood were also reported during this period (Table 2).

All the hospital discharges for V. parahaemolyticus infection were coded to A053 (foodborne V. parahaemolyticus intoxication), although it is not clear whether foodborne transmission was suspected, or the code was used because it is the only specific code for V. parahaemolyticus (mixed culture wound infections were evident in three cases).

Recreational contact with coastal or river water in New Zealand was the implicated cause of five cases of V. parahaemolyticus infection reported to EpiSurv (two cases appear to form an unidentified household cluster). One additional case may have been occupationally exposed through processing fish. Two cases reported contact with sick people but both cases also consumed foods that were implicated as the source of their infection. Drinking water was not an implicated source for any of the V. parahaemolyticus cases.

Discussion

Vibriosis and cholera cases are recorded in different New Zealand health databases that support the national notifiable disease system but have different purposes. Through examining these data, a picture emerges of infections involving various Vibrio spp., primarily manifesting as gastrointestinal, soft tissue or ear infections, sometimes with severe health outcomes. We have described some of the species-specific, regional exposure and seasonal trends, which reflect what is observed elsewhere.¹ It is apparent that nationally there are unquantified numbers of soft tissue infections and sepsis due to V. vulnificus and V. parahaemolyticus, gastrointestinal illness due to V. vulnificus and non-toxigenic V. cholerae and ear infections involving Vibrio spp. The data on V. alginolyticus infections suggest a common pattern of ear infections among young males, likely arising from recreational exposure to seawater.⁷ The ERL data show that people are being exposed to non-O1/non-O139 V. cholerae within New Zealand, with resultant illness, and potential sources should be a focus of future research. Almost one-third of the hospital discharge records, and the only three vibrio-related deaths in the mortality database, included the coding “cholera, unspecified”. It is not known whether these records related to infections from non-toxigenic V. cholerae, but the findings signal the need for deeper investigation.

The hospital discharge data show that V. vulnificus is a cause of serious infections in New Zealand, particularly among males. These data highlight diabetes mellitus and hepatitis B as comorbidities. Haemochromatosis, liver disease including alcoholic cirrhosis, diabetes mellitus and renal disease are recognised risk factors for V. vulnificus infection.²² In New Zealand, only one fatality has been reported associated with V. vulnificus infection, in 1990, as a result of a wound infection.²³ In 2008, there was a single New Zealand case of fatal necrotising fasciitis and associated fulminant sepsis in a 79-year-old male, attributed to exposure of a skin wound to V. parahaemolyticus in seawater.²⁴ This fatality did not emerge among the hospital mortality data, most likely because this was not a foodborne case and thus did not meet the description of the V. parahaemolyticus ICD10 A05.3 code.

There are likely to be people groups more at risk for infection or severe health outcomes due to their daily activities and health status. Pacific peoples and Māori appear to be over-represented among vibriosis cases. However, further study is needed to identify the factors that increase the likelihood of any individual becoming infected by Vibrio spp., with ethnicity data only providing some direction for these enquiries.²⁵ Case-specific information about food and environmental exposures (including their location and timing) is required to better understand domestically acquired vibriosis in New Zealand, and is vital for risk management. For non-notifiable vibriosis cases, exposure assumptions were made in this study. For example, the hospital discharge data show cases of V. vulnificus wound infection and sepsis, but this does not prove the source was environmental (marine or estuarine). While case exposure information is lacking, this current analysis does show that recreational exposure to aquatic environments and consumption of seafood (particularly shellfish) are both important transmission routes in New Zealand, which is in keeping with the ecology of non-cholera Vibrio spp.

Temporal trends are affected by changes to data collection, laboratory protocols and outbreak events, which mask any true changes to vibriosis prevalence over time. Cases of V. parahaemolyticus AG were specifically listed as being notifiable under the disease classification “acute gastroenteritis” from 2012.¹⁵ Thus, cases may have been notified as AG without listing the causal pathogen, and therefore not captured in the EpiSurv data extract used for this study.¹⁴ Conversely, the V. parahaemolyticus outbreaks occurring from 2019 increased national efforts towards understanding and minimising foodborne exposure, resulting in improved case follow-up and recording of information in EpiSurv. The V. vulnificus–specific ICD code (B96.82) was introduced for use in New Zealand from 2014. Information collected in 2005 shows that only a small number (two out of 13) of New Zealand diagnostic laboratories routinely tested faecal specimens for Vibrio spp.²⁶ Since 2015, the increased use of culture-independent diagnostic testing of faecal samples will have increased Vibrio spp. detection.²⁷

New Zealand’s current national communicable disease reporting system captures cases of notifiable diseases as defined by the Health Act 1956.¹⁵ It is not designed to demonstrate the breadth of vibriosis occurring in New Zealand. The decentralised way that Vibrio spp. infections are currently recorded limits the ability to measure the national extent of morbidity and mortality from domestically acquired infections and to provide robust evidence that could underpin public health interventions in higher-risk regions, during higher-risk periods or among at-risk people groups. It is also not currently possible to provide an objective baseline to measure or predict future epidemiological changes in response to local short- or long-term environmental events.

In the United States of America, vibriosis has been nationally notifiable since 2007 and all cases are reported to a dedicated system (COVIS).²⁸ The most recent data available (2019) show that COVIS captures infections from a range of Vibrio spp., including those rarely reported as human pathogens.²⁹ COVIS has enabled recognition of changes in vibriosis incidence over time and triggered further investigations.^10,30 This surveillance system is a useful model to consider for New Zealand. Surveillance that captures all vibriosis and associated metadata would provide valuable epidemiological information and a baseline to monitor or predict any changes. Such a system could direct risk management activities and support health service planning and resilience.