A model for rapid, active surveillance for medically-attended acute gastroenteritis within an integrated health care delivery system

Abstract

Background: This study presents a novel methodology for estimating all-age, population-based incidence rates of norovirus and other pathogens that contribute to acute gastroenteritis in the United States using an integrated healthcare delivery system as a surveillance platform.

Methods: All cases of medically attended acute gastroenteritis within the delivery system were identified from April 1, 2014 through September 30, 2016. A sample of these eligible patients were selected to participate in two phone-based surveys and to self-collect a stool sample for laboratory testing. To ascertain household transmission patterns, information on household members with acute gastroenteritis was gathered from participants, and symptomatic household members were contacted to participate in a survey and provide stool sample as well.

Results: 54% of individuals who met enrollment criteria agreed to participate, and 76% of those individuals returned a stool sample. Among household members, 85% of eligible individuals agreed to participate, and 68% of those returned a stool sample. Participant demographics were similar to those of the eligible population, although minority racial/ethnic groups were somewhat underrepresented in the final sample.

Conclusions: This study demonstrates the feasibility of conducting acute infectious disease research within an integrated health care delivery system. The surveillance, sampling, recruitment, and data collection methods described here are broadly applicable to conduct baseline and epidemiological assessments, as well as for other research requiring representative samples of stool specimens.

Fig 1. Overview of MAAGE surveillance design and process.

^aMAAGE: Medically-attended acute gastroenteritis. ^bOSPHL: Oregon State public health lab. ^cHHM: Household member.

Fig 2. Stool sample collection kit provided to enrolled MAAGE study participants.

Stool sample collection kit and information sheet (kit provided by OHPL)

1. • Screw-top plastic container (for the specimen). Collection container will have a label for Health Record Number (HRN), Name, date of birth (DOB), date of collection, time of collection. Recruitment staff will include member’s HRN, Name and DOB before sending kit to participant.

2. • Card board-and-tissue-paper liner (fits on a toilet seat) with paper bowl (can be added to liner if needed)

3. • Spoon (to scoop)

4. • Gloves

5. • Gauze pad (for use with diapers)

6. • Alcohol cleaning pad

7. • Plastic specimen bag (may say "Biohazard") with absorbent towel

8. • Paper sack

Refrigeration Kit

1. • Cold pack(s)

2. • Chill Checker button (count of 500 for randomly selected samples)

3. • Information sheet for keeping specimen at required temperature

4. • Sending/Returning shipping packet and information sheet

5. • Shipping label for return sample

6. • Shipping box

Fig 3. KPNW members identified with medically-attended acute gastroenteritis, April 1, 2014-September 30, 2016.

^aParticipants with a single AGE encounter per day, prior to applying exclusion criteria. ^bEncounters occuring ≥30 days following the preceding encounter. ^cHealth plan member with ≥1 AGE encounter during the study period who meets enrollment criteria; members can have multiple events. ^dAll recruitable members aged <5 and ≥75 years of age; 35% of the following age groups: 5–17, 18–44, 45–64, 65–74. ^eContacted and meets all criteria for study population. ^fAgreed to participate and completed the baseline survey. ^gReturned samples were not pathogen-tested; 8 were rejected by the KPNW lab due to inadequate information; 49 were deemed nonviable by the Oregon State Public Health Lab.

Fig 4. KPNW household members of enrolled MAAGE participants, April 1, 2014-September 30, 2016.

^aRecruited household members of participants that have onset of AGE symptoms ≤7 days before primary participant’s recruitment or follow-up call. ^bAgreed to participate and completed baseline survey. ^cReturned samples were not viable for pathogen testing.

Fig 5. Specimen return characteristics from KPNW members enrolled in MAAGE.

*Missing values (n = 8, n = 4) due to untestable stool samples. ¹Encounter: Date of patient’s first medically-attended acute gastroenteritis (MAAGE) encounter with Kaiser Permanente Northwest (KPNW) healthcare system. ²Collection: Date stool sample was collected by participant. ³Receipt: Date sample was received by study staff at KPNW.