An environmental investigation was conducted to evaluate the potential for spring water to contribute to the continued transmission of HAV in the community. This investigation consisted of site inspections, a sanitary survey, and drinking water and wastewater sample collection. ACHD initially visited the property in March of 2019 for sampling and the on-site environmental assessment, which included interviews with the property owner and assessment of the spring water location, flow, and patterns. Observations and inspection findings were documented on an environmental health outbreak investigation survey data collection tool. ACHD and the Kentucky Division of Water evaluated the resident’s black water tank used for all human fecal waste from the home for possible leaks in the system via a charcoal smoke test (National Association of Sewer Service Companies 2021). In December 2019, the property owners and downstream farms requested that ACHD retest the spring so they could commence using the spring water for irrigating produce crops sold locally. The ACHD conducted a subsequent site visit in January of 2020 for resampling water and wastewater to evaluate the persistence of viral shedding from the household and environmental contamination.

Environmental sampling: A map of the water sources and sampling sites is provided in Figure 2. On 11 March 2019, ACHD collected large volume (100-L) water samples through dead-end ultrafiltration (DEUF) (Gallardo et al. 2022) at two sites on the property for HAV testing (Jothikumar et al. 2005): (1) the site of drinking water collection from the main spring; and (2) immediately downstream of the home spring where the household gray water (i.e., non-human waste spent household water) discharged from a pipe (referred to as gray water downstream in Figure 2). The team also collected two small-volume (100-mL) water samples from each DEUF collection site for fecal indicator organism (i.e., total coliforms, Escherichia coli (E. coli), and somatic coliphage) testing (Gallardo et al. 2022). A 100-mL grab sample was collected from a drinking well that served the community school for fecal indicator bacteria (FIB) (total coliforms and E. coli only) testing. The school groundwater well ~3 km (<2 mi) from the household where cases were identified was sampled due to concerns from the community that if the springs were contaminated there could be larger aquifer contamination; among outbreak cases, one case of illness was confirmed by laboratory testing for a schoolteacher and epidemiological links established for two students, one of which attended the wedding.

On 25 March 2019,ACHD collected five additional large volume DEUF samples for HAV testing from: (1) the three upstream spring feeding the home main spring; (2) the black water storage tank at the home; and (3) further downstream of the gray water discharge pipe. Four 100 mL grab samples were also collected from a nearby roadside spring, neighbor spring, secondary spring, and pond hydraulically feeding into the home main spring for somatic coliphage and FIB testing (Figure 2).

On 6 January 2020, ACHD conducted resampling for HAV using DEUF from three of the 2019 sampling sites: (1) the home main spring; (2) immediately downstream of the main spring and gray water discharge pipe; (Figure 2, referred to as gray water downstream); and (3) a site further downstream of the gray water discharge pipe near the neighbor’s spring. These sites were selected because the previously affected household and downstream farms wanted to resume spring water use for irrigating fresh produce crops. Paired small-volume (100-mL) water samples were also collected and tested for FIB and somatic coliphage, like the 2019 samples.

In late 2019, the black water tank was pumped out and filled with building materials to prevent reuse, and a new septic system was installed. On 17 February 2020, two wastewater samples (950 mL and 1 L) were collected from the household’s new septic tank for HAV and fecal indicator testing. Fecal indicator (generic and human-specific) testing of the septic tank was conducted to evaluate the abundance of these indicators within the small, on-site waste system to determine the validity in using these indicators for tracking human fecal waste from the system to nearby springs.

Water and wastewater samples were transported overnight to the CDC in Atlanta, GA, on ice for processing within 24 h of collection. At each sampling site, ACHD collected water temperature and pH using a Hach Multimeter Pocket Pro +, except for the school well and surrounding springs.

In the initial 2019 environmental sampling, CDC tested DEUF water samples for HAV, total coliforms, E. coli, and the general viral fecal indicator, somatic coliphage. For the follow-up 2020 environmental testing, DEUF spring water and wastewater grab samples were tested for the 2019 microbes and also tested for the molecular-based human-specific viral fecal indicator, crAssphage (Stachler et al. 2017), to better track household human waste contamination to the spring. For all sampling events, total coliforms and E. coli were enumerated in 100-mL grab samples via the IDEXX QuantiTray-2000 method using Colilert-18 media (Standard Methods Committee 2018). Turbidity was also measured in grab samples from the four surrounding spring water samples using a Hach 2100P Turbidity Meter in nephelometric turbidity units (NTU).

DEUF samples were processed as previously described (Yard et al. 2014). Briefly, ultrafilters were first backflushed (~700 mL backflush concentrate volume). A portion of the DEUF backflush (20 mL) was tested directly for somatic coliphage using a modified version of US Environmental Protection Agency (USEPA) Method 1602 (U.S. Environmental Protection Agency 2001). The coliphage method was modified so that each plate contained 5 mL of sample in 30 mL of 1× media with the same concentrations of appropriate antibiotic, host, and MgCl₂ in the final volume as Method 1602. Culture positive controls and media blanks were run in parallel to all samples. The following control strains were used for the culture-based microbial assays: 30 colony-forming unit (CFU) BioBalls^® (BTF Pty. Ltd, USA) for coliforms and E. coli (National Collection of Type Cultures 9001), and bacteriophage-ɸX174 (ATCC 13706-B1) for somatic coliphage (ATCC, Manassas, VA).

The remaining DEUF backflush was then subjected to secondary concentration by polyethylene glycol (PEG) precipitation and centrifugation (~3 mL final concentrate volume) as previously described (Yard et al. 2014). Briefly, PEG precipitation was conducted using 8% (wt./vol) PEG and 0.3 M (wt./vol) NaCl to the remaining backflush volume, incubated at room temperature for at least 2 h, and then repeatedly centrifuged in polycarbonate 250 mL bottles at 10,000 × g for 30 min at 4 °C. Supernatant was then removed and more sample added until all sample volume was concentrated. A portion of this concentrate (750 μL) was added to bead beating tubes and stored at 4 °C until nucleic acid extraction and testing for molecular targets (HAV and crAssphage). The remaining portion of the concentrate was also tested for somatic coliphage as described earlier, and volumes with countable plates were reported.

DEUF water sample concentrates (750 μL) were extracted for total nucleic acid as previously described with a final extraction elution volume of 80 μL for water samples and 50 μL for septic (Kahler et al. 2025). An extraction blank containing only reagents was run alongside the samples extracted from each sampling event. Extracts were stored for 24 h at 4 °C until molecular testing and stored at −80 °C until genomic analysis (if applicable).

HAV RNA was detected in water concentrates using reverse-transcriptase quantitative polymerase chain reaction (RT-qPCR) as described in Jothikumar et al. (2005). CrAssphage DNA (056 assay) was detected using qPCR as described in USEPA Method 1696 (U.S. Environmental Protection Agency 2019). Each sample was analyzed in triplicate with either 5 μL of extract in 50-μL reactions (RT-qPCR) or 2 μL of extract in 25-μL reactions (qPCR) and using either Taqman^™ Fast Virus 1-Step Master Mix (ABI) (RT-qPCR) or Environmental MasterMix 2.0 [Applied Biosystems (ABI), Waltham, MA]) (qPCR), respectively, on an Applied Biosystems 7500 thermocycler (ABI). To test for amplification inhibition, each sample template was tested at a 1:10 dilution, and the presence of inhibition was defined by less than a 2.3 difference between C_q values for the diluted and undiluted samples (Boehm et al. 2013).

For each instrument run, the amplification threshold was set to 0.03 ΔRn. Three no-template controls (NTC) were included with each instrument run to evaluate reagent contamination of the target. A (RT-)qPCR result was considered a detection if at least two reactions amplified at a quantification cycle (C_q) value < 40 or if any amplified product was confirmed to be the target via sequencing. Otherwise, the result was considered a non-detect. CrAssphage efficiency and sensitivity were determined using standardized linearized synthetic DNA reference material (NIST SRM^® 2917) consisting of five 10-fold serial dilutions (10¹ to 10⁵ copies per reaction) (Willis et al. 2022). CrAssphage concentrations were determined using the standard curve equation and volumes of sample processed and analyzed.

HAV detected in water samples underwent genomic analyses using targeted amplicon sequencing by the CDC Division of Viral Hepatitis laboratory. A 349-nucleotide segment of the HAV genome spanning the VP1–P2B junction was reverse transcribed, and the cDNA was PCR amplified and subjected to next-generation sequencing using the CDC-developed global hepatitis outbreak and surveillance technology (GHOST) protocol (Ramachandran et al. 2021). Genotyping and phylogenetic analysis using Neighbor joining algorithm and Kimura-80 Distance (CLC Genomics, Workbench v24, Qiagen, Aarhus) were conducted to determine the relatedness of the sequence to RNA-positive surveillance and outbreak samples collected in the US during the HAV outbreaks of 2016–2020 (Ramachandran et al. 2021; Foster et al. 2022).

The property where the wedding was held contained two homes; the second household was abandoned when the family moved away in December 2018, before the wedding. The spring and water collection site was located, 8 m (25 feet) down gradient from the primary home. The property residents utilized the main spring for household water use, including drinking. The spring water was reportedly transported and stored in coolers and tanks for home use. The spring water was also used to irrigate produce sold to the local community. After the initial outbreak during December 2018, the 22,000-L (7,000-gallon) drinking water storage tank was reportedly manually treated with chlorine in 2019.

The single outhouse used by the household where hepatitis A cases were identified was connected to a black water tank waste system. The black water storage tank was reportedly pumped out periodically by a separate company (frequency not reported). The charcoal smoke test did not detect leaks in the system. Gray water from the home was piped out into a stream created by the drinking water spring about 55 m (180 feet) downstream from the drinking water collection point. A large amount of heavy rainfall had been reported in the area during that winter.

In March of 2019, HAV was detected by molecular-based methods (RT-qPCR) in both the main spring water and in the downstream water of the household gray water discharge pipe (both immediately and further downstream) (Table 1). The spring water had 1,203.3 most probable number (MPN) per 100 mL total coliforms and 58.3 MPN/100 mL E. coli. The water downstream of the household gray water pipe had 1,203.3 MPN/100 mL total coliforms and 102.2 MPN/100 mL E. coli. No FIBs were detected in the school groundwater well sample. HAV was also detected in the storage tank at the home. HAV was not detected in any of the three springs (Figure 2: Upstream Spring 1, 2, and 3) feeding the main spring, but somatic coliphage were detected in all three springs and the main spring (Upstream Spring 1, 2, and 3: 5.53, 2.21, and 5.2 plaque forming units/100 mL, respectively; main spring: 2.5 PFU/100 mL). Total coliforms, E. coli, and coliphage were also detected in the surrounding waters (Figure 2: roadside spring, neighbor spring, and secondary spring on the property).

The sequence obtained from the spring water HAV detection covers a region spanning the VP1/P2B junction region of the HAV genome and identifies the detected HAV as genotype IB. The sequence was identical to a cluster of outbreak sequences obtained from Kentucky and very close to other unique genotype IB strains identified via surveillance or outbreak investigations in the United States throughout the period 2016–2020 (Ramachandran et al. 2021; Foster et al. 2022), as shown in Figure 3. The leaves of the tree represent either single isolates or clusters (several cases with identical sequences); however, the frequency of each unique strain is not shown.

From the 1-year follow-up testing in January of 2020, HAV was detected in the main spring and gray water downstream from the main spring (Table 2). Total coliforms and E. coli were detected in the main spring at levels above EPA drinking water standards (,1 E. coli per 100 mL) (2,419.6 and 58.5 MPN/100 mL, respectively). Somatic coliphage, viral fecal indicators, were also detected at all three sampling sites (main spring = 14.26 PFU/100 mL; downstream = 12.31 PFU/100 mL; neighbor spring = 25.45 PFU/100 mL), confirming viral-associated fecal contamination. The human-specific viral fecal indicator, crAssphage, was also detected in all three water samples; the concentration decreased farther downstream (Table 2).

In February of 2020, HAV was detected in the two small-volume wastewater samples collected from the newly installed, 3,800-L (1,000-gallon) septic tank (Table 2). Total coliforms and E. coli were at expected levels in the septic wastewater (total coliforms: 9.2–9.8 × 10⁶ MPN/100 mL; E. coli: 5.8–11.4 × 10⁵ MPN/100 mL), suggesting typical fecal input into the septic system. Somatic coliphage, viral fecal indicators, were also detected at expected levels in the septic wastewater (1.8–2.3 × 10³ PFU/100 mL). Finally, the human-specific viral fecal indicator, crAssphage, was detected in the septic wastewater samples (1.25 × 10⁶ copies/100 mL). Together, these data suggest that both the general and human-specific phages were indicators of septic wastewater from this household.

No contamination was observed in any culture-based indicator assays, nucleic extractions blanks, or NTC. Positive controls were detected as expected for all assays. (RT-)qPCR amplification inhibition was absent from all samples. The crAssphage pooled standard curve (y = −3.5347 x + 42.612) had an amplification efficiency (E) of 0.9183 (E = 10^{(−1/slope−1)}) and R² of 0.9967. The lower limit of quantification of the crAssphage assay had an average C_q of 38.95 (1 log10 copies per reaction).