Cross-sectional investigation of household transmission of Cryptosporidium in England and Wales: the epiCrypt study protocol

Introduction Infection with the Cryptosporidium parasite causes over 4000 cases of diagnosed illness (cryptosporidiosis) in England and Wales each year. Risk factors are often estimated from outbreak investigations, and in the UK include ingestion of contaminated water and food, farm/animal contact and person-to-person spread in institutions. However, reported outbreaks only represent about 10% of cases and the transmission routes for sporadic disease may not be the same. Contact with other people has been highlighted as a factor in the transmission of Cryptosporidium, but the incidence of sporadic disease has not been sufficiently established, and how frequently this arises from contact with other infected people is not well documented. This project will estimate the amount of secondary spread that occurs in the home and potentially identify asymptomatic infections which might have a role in transmission. Risk factors and characteristics associated with secondary spread will be described including any differences in transmission between Cryptosporidium species. Methods and analysis The study will prospectively identify cryptosporidiosis cases from North West England and Wales over 1 year and invite them and their household to take part. Each household will complete a questionnaire and each household member will be asked to provide a stool sample. Clinical, demographic and home variables will be described, and further analyses undertaken to investigate associations with secondary spread in the home. Cryptosporidium-positive stool samples, identified by immunofluorescence microscopy, will be characterised using molecular methods to describe patterns of transmission. Data collection is expected to take 1 year, beginning in September 2018. Ethics and dissemination The study has been approved by the North West–Liverpool East NHS Research Ethics Committee (Reference: 18/NW/0300) and the Confidentiality and Advisory Group (Reference 18/CAG/0084). Outputs will include scientific conferences and peer-reviewed publications. In addition, a short, lay report of findings will be produced for participants, who can opt to receive this when they take part. Trial registration number CPMS ID: 39458.

probably propagated by person-to-person spread in the home [7]. An analysis of surveillance 135 data in Ireland reported that, acknowledging the number of community outbreaks reported in 136 Ireland, it is likely that indigenous cases form the majority of illness and that among these 137 ingestion of water and person-to-person spread are the most important mechanisms of 138 transmission [8]. In the United States, a study evaluated sporadic cryptosporidiosis among 139 immunocompetent persons using a case-control design. Risk factors associated with 140 increased odds of being a case were international travel, contact with cattle, and contact with 141 a child with diarrhoea [13]. In 2001-02, a case-control study conducted in the North West of 142 England examined species-specific risk factors for sporadic cryptosporidiosis [9]. The 143 authors compared risk factors for infection with genotypes 1 and 2 (currently recognised as 144 C. hominis and C. parvum respectively) and found that contact with another person with 145 diarrhoea was a risk factor for infection with Cryptosporidium, and that exposure through 146 changing children's nappies was a specific risk factor for infection with C. hominis whether 147 the child was symptomatic or not. 148 A Norwegian study looking at follow-on spread after two outbreaks where children were  Cryptosporidium, then sporadic cases may subsequently arise following exposure to either, 162 and outbreaks in close settings such as the home may happen more frequently than is 163 currently recognised. Few studies exist which refute or confirm this, especially in 164 industrialised countries. 165

AIMS AND OBJECTIVES 166
The aim of this study is to estimate the amount of secondary spread that happens in the 167 home and describe associated factors and case characteristics. This study will support our 168 Cryptosporidium isolates were reported from the North West (8.4/100,000 population) [18]. 188 Wales has a total population of over 3 million people [19]. In 2016, over 400 laboratory-189 confirmed Cryptosporidium isolates were reported from Wales, the highest rate of 190 Cryptosporidium spp laboratory reports per 100,000 population in England and Wales 191 (15/100,000) [18]. 192

Surveillance/sampling frame 193
The sampling frame will be taken from the two relevant surveillance systems which capture All cases of laboratory confirmed Cryptosporidium sp. reported from primary diagnostic 203 microbiology laboratories in North West England and Wales, in the study year, will initially be 204 eligible. 205

Study type 206
The identification of cases, and their subsequent recruitment, is cross-sectional, although the 207 study also involves retrospective data collection and some prospective sampling (Error! 208 household members (excluding the index case) will be asked to supply a stool sample. 213

Study period 214
The study period will be 12 months, to account for seasonal variation and allow maximum 215 enrolment, up to 400 households. The study is expected to begin with a pilot phase of 1-2 216 months in autumn 2018. Tables 1 and 2 outline case and household definitions used to categorise household members and define secondary transmission. 236

Case definitions Index case
The first case from a household identified in the surveillance system (person reported to a PHE/PHW surveillance system(s) following detection of Cryptosporidium sp. in a faecal sample, with a specimen date in the study year)

Secondary case(s)
Probable secondary case A person in a household of an index case, with symptoms: of diarrhoea and/or vomiting AND that started after another case's onset date in the household Confirmed secondary case A person in a household of an index case, with symptoms: of diarrhoea and/or vomiting AND That started after another case's onset date in the household; AND a Cryptosporidium positive stool sample Asymptomatic A person in a household of an index case with: no reports of gastrointestinal illness; AND a Cryptosporidium positive stool sample.  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47 o n l y

Household definitions Household
Two or more people (not necessarily related) living at the same address in North West England or Wales who share cooking facilities and share a living room or sitting room or dining area [18].

Household member
A person who normally resides in the household and regularly shares food or toilet facilities [26] Household contact A household member where an index case has been identified Household with transmission A household that has more than one case Household without transmission A household that has one case (the index case)

Recruitment approach 239
Overview and rationale 240 Cases of cryptosporidiosis are identified from routine surveillance (from diagnostic 241 laboratories) and are contacted via post, by the relevant public health organisation, in the 242 first instance. Following this, if they do not opt-out, they are contacted via telephone by an 243 NHS research nurse at the local Clinical Research Network (CRN) to chat about the study 244 and determine if they would like to take part. 245 Our approach to the recruitment process was driven by necessity and feasibility and we 246 explored several options at the protocol drafting stage of the project, balancing data needs 247 with patient choice. As detection of Cryptosporidium is retrospective and undertaken by 248 laboratory staff, there is no opportunity to consent individuals at the time of diagnosis and 249 the recruitment process could not be achieved without access to patient information. In our 250 model, participants are given opt-out options at each contact and it is emphasised that they 251 can withdraw at any time. Previous research supports the acceptability and understanding of 252 this method, recognising that an approach of 'consent for each use' is burdensome for both 253 researcher and participant [27] [28], as does patient response and engagement with similar 254 studies. 1 255

Public and Patient Involvement 256
Patients and public were not involved in the overall design of the study, but we did elicit 257 some public opinion when finalising our approach to recruitment. Following valuable 258 comments from the ethical review board we undertook a short survey among the public and 259 specific Patient and Public Involvement (PPI) groups to gauge general attitudes toward 260 accessing data prior to consent, to support recruitment to research. We drafted a survey 261 which outlined the approach to recruitment and the framework of the study. We accessed a 262 lay PPI group from the Infection and Global Heath panel at the University of Liverpool, and 263 the method of recruitment and how they felt about this approach. In general, the feeling was 265 that it is acceptable to access data for recruitment, especially to support much needed 266 research. However, considerations and worries included the person accessing data, with 267 NHS/public health staff generally viewed as more favourable that non-NHS (e.g. University) 268

Identification and first contact with the index case 270
Laboratory diagnosed reports of Cryptosporidium, and the corresponding patient contact 271 details, will be extracted from the relevant surveillance system by health protection staff and 272 saved in a line list. 273 All potentially eligible participants will be issued a unique sequential study ID by PHE/PHW 274 staff in the initial line listing. This will be on all relevant study documentation and stool pots 275 and follow each person and household through the study journey. This will allow data to be 276 linked pseudonymously and helps with data management. 277 Staff from either PHE or PHW (depending on case location) will send an invite letter through 278 the post to these potentially eligible index cases. The invite letter outlines the study, 279 describes why the case has been contacted, and explains that a research nurse may be in 280 touch over the coming weeks to discuss the study. The letter allows the case to choose 281 several ways of opting-out of this contact (e-mail, freepost, telephone) and provides a 282 named, clinical study lead for each public health organisation should they wish to discuss 283 any aspect of this. 284

Approaching to recruit 285
If a contacted case does not opt-out within two weeks, their details will be shared securely 286 (using internally agreed practices) with the NHS research nurses at the Clinical Research 287 Network North West Coast (CRN). The research nurses will attempt to contact the index 288 case (or parent/guardian of,) via telephone (using internally agreed practices) to inform them 289 about the study and offer them the opportunity to participate, if eligible. A maximum of three 290 F o r p e e r r e v i e w o n l y attempts will be made, and nurses will not leave voicemails. If a case is unable to be 291 contacted, or does not wish to participate at this stage, their details will be deleted from the 292 line list. If the approached index case is successfully contacted via telephone and interested 293 in participating, or would like more information, the research nurses will prepare and post a 294 study pack. Index cases may be excluded at this stage where discussions with the case 295 reveal that any of the following exclusion criteria apply; 296 pseudonymised for analyses. Name, date of birth, and full postcode will be removed and 344 replaced with unique study ID, age, and Lower Super Output Area (LSOA). 345 All data storage, cleaning and analyses will be undertaken by the study team at the 346 University of Liverpool. Data will be stored on institutional network drives with appropriate 347 security measures in place. Hard copy records will be stored in a locked cabinet in a secure 348 location and access to records and data are limited to study personnel. Paper documents 349 will be stored separately from the corresponding electronic data. The sponsor and data 350 controller for this study is the Clinical Governance Team at the University of Liverpool. 351

Stool sample management 353
For purposes of data confidentiality and governance, stool samples returned to the 354 Cryptosporidium Reference Unit are pseudonymised with the unique study ID, and 355 participants will be asked to write their age, sex, and date of sample on their sample pot 356 before they collect the stool. Stool sample results will be added to a study-specific database 357 held at the CRU. The stool sample results will be added to the study database at the end of 358 data collection using unique study ID in a secure file transfer. 359

Questionnaire data 368
The questionnaire is divided into sections and is mostly composed of dichotomous and 369 multiple-choice questions. Section A asks questions to determine the composition of the 370 household, the clinical details of the index case, and captures any other symptomatic 371 household members. A table is used to collect information on any other symptomatic 372 diarrhoeal illness in the house and will attempt to capture relationships to the index case 373 (  Section B records activities of the index case, and others in the home, in the two weeks prior 377 to the index case's onset, based on known exposures for Cryptosporidium. Information on 378 outdoor and leisure activities may help determine possible co-primary infections and 379 distinguish them from those that are secondary. (Table 4) 380 Sections C and D collect household variables, including the number of bedrooms and 381 bathrooms, and capturing those who share beds/baths, and asking about outside space and 382  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  Visiting or working on a farm or had contact with farm animals Visiting or working at a zoo or had contact with zoo or wild animals  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  60   F  o  r  p  e  e  r  r  e  v  i  e  w  o n l y

Stool collection and genotyping 394
All household members of the index case (but not the index case) will be asked to provide a 395 stool sample using the Fe-Col® kit provided in the study pack, and post to the 396 Cryptosporidium Reference Unit. (Figure 3) 397 The stool pots will be labelled with the unique household ID which identifies them as part of 398 the study but allows the samples to remain anonymous to the reference laboratory team. 399 Samples will be scored against the Bristol stool scale and tested, only for Cryptosporidium, 400 using Immunofluorescence (IF) microscopy. Positive samples will be further processed by 401 oocyst purification and DNA extraction and then Cryptosporidium speciated using validated 402 PCR techniques which are part of normal practice [30]. Cryptosporidium DNA will be 403 retained for subtyping and possibly whole genome sequencing (WGS) at a later date. 404 Full laboratory protocol available on request 405 406

ANALYSES 407
The primary objective of this study is to determine the amount of secondary spread that 408 happens in the home following exposure to a case. This will be established by testing stool 409 samples of household members of a case for Cryptosporidium and reporting the numbers of 410 secondary cases (according to our pre-determined definitions). A household with more than 411 one case will be a household with secondary transmission. 412 We will calculate the following: 413  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  Confounding (e.g. host factors such as age, co-morbidity) will be considered, where known, 423 using multivariable techniques. Also, we will, where possible, examine environmental level 424 exposures using stratification, e.g. those households/cases which are exposed to other 425 known sources or risk factors, such as those living on farms. 426 Data will be analysed, where appropriate, using Stata v12. 427

LIMITATIONS AND BIASES 429
Some elements of the study design are retrospective in nature, as the index case must have 430 already been ill and been tested in order to be selected. As a result, some ascertainment 431 bias may lead to a skewed sample from which to choose the index cases. We do not expect 432 to capture the full profile of cases and households in the population that might have 433 Cryptosporidium due to differences in risk or vulnerability, severity, or health-seeking 434 behaviour[31]. As Cryptosporidium is common in younger age groups we expect a large 435 proportion of the participants to represent families with young children. Any likely over-or 436 under-representation in the data collected will be considered when assessing and describing 437 results. Further unidentifiable limitations may include recall biases around dates of onset or 438 activities, and classification biases as we are asking about self-reported illness and 439 information may be inaccurate. 440 There is a possibility that we could see ongoing outbreaks in the study year. If this happens, 441 we might try to identify these where possible and may consider excluding households from 442 this study where all or most members have been exposed, including a definition of aco-443 primary case.  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  60   F  o  r  p  e  e  r  r  e  v  i  e  w  o n l y

End of study 446
The study will be declared as ended when the database is closed to recruitment -after one 447 year or when the maximum number of households has been enrolled. 448 449

Competing interests statement 576
The authors declare that they have no competing interests.

Other information
Funding 22 Give the source of funding and the role of the funders for the present study and, if applicable, for the original study on which the present article is based *Give information separately for cases and controls in case-control studies and, if applicable, for exposed and unexposed groups in cohort and cross-sectional studies.

77
This project will estimate the amount of secondary spread that occurs in the home and 78 potentially identify asymptomatic infections which might have a role in transmission. Risk 79 factors and characteristics associated with secondary spread will be described including any 80 differences in transmission between Cryptosporidium species.

81
Methods and analysis

82
The study will prospectively identify cryptosporidiosis cases from North West England and

83
Wales over one year and invite them and their household to take part. Each household will 84 complete a questionnaire and each household member will be asked to provide a stool 85 sample. Clinical, demographic, and home variables will be described, and further analyses 86 undertaken to investigate associations with secondary spread in the home.

87
Cryptosporidium-positive stool samples, identified by immunofluorescence microscopy, will 88 be characterised using molecular methods to describe patterns of transmission. Data 89 collection is expected to take one year, beginning in September 2018.   cases may subsequently arise following exposure to either, and outbreaks in close settings 182 such as the home or institutions may happen more frequently than is currently recognised.

183
Few studies exist which refute or confirm this, especially in industrialised countries.

185
The aim of this study is to estimate the amount of onward spread of Cryptosporidium that 186 happens in the home, and to describe associated factors and case characteristics. ( The identification of cases, and their subsequent recruitment, is cross-sectional, although the 237 study also involves retrospective data collection and some prospective sampling (Error!

239
Index cases of cryptosporidiosis will be identified via the relevant surveillance system(s).

240
Once participants are recruited, they will complete a questionnaire (one per household), The first case from a household identified in the surveillance system (person reported to a PHE/PHW surveillance system(s) following detection of Cryptosporidium sp. in a faecal sample, with a specimen date in the study year) Secondary case(s) 1 Probable secondary case A person in a household of an index case, with symptoms: of diarrhoea and/or vomiting AND that started after another case's onset date in the household Confirmed secondary case A person in a household of an index case, with symptoms: of diarrhoea and/or vomiting AND That started after another case's onset date in the household; AND a Cryptosporidium positive stool sample Asymptomatic A person in a household of an index case with: no reports of gastrointestinal illness; AND a Cryptosporidium positive stool sample. 1 We use the term 'secondary spread' to mean any apparent onward transmission of disease originating from an index case, whilst recognising that this may be secondary or even tertiary levels of spread

Household definitions Household
Two or more people (not necessarily related) living at the same address in North West England or Wales who share cooking facilities and share a living room or sitting room or dining area [18].

Household member
A person who normally resides in the household and regularly shares food or toilet facilities [32]

Household contact
A household member where an index case has been identified Household with transmission A household that has more than one case Household without transmission A household that has one case (the index case) 278  Visiting or working on a farm or had contact with farm animals Visiting or working at a zoo or had contact with zoo or wild animals 430 All consenting household members of the index case (but not the index case) will be asked 434 to provide a stool sample using the Fe-Col® kit (Figure 3) provided in the study pack, and 435 post to the Cryptosporidium Reference Unit.

436
The stool pots will be labelled with the unique household ID which identifies them as part of 437 the study but allows the samples to remain anonymous to the reference laboratory team. . We may get an over-representation of severe disease as these cases are 474 more likely to seek health care and be tested, and perhaps more likely to test positive.

475
We are only collecting one sample from each household member, and not re-sampling the 476 index case, for time and resource reasons. This may well lead to missing intermittent 477 shedding of oocysts, tertiary household infections and/or misclassifying recurring illness.

478
As Cryptosporidium is common in younger age groups, we expect a large proportion of the 479 participants to represent families with young children which may lead to over-representation 480 of these households. In addition, we might expect that having young children who were ill, or 481 being severely ill themselves, may incentivise cases to participate in the study, more than 482 adult, less severe cases.

483
Any likely over-or under-representation in the data collected will be considered when 484 assessing and describing results. Further unidentifiable limitations may include recall biases 485 around dates of onset or activities, and classification biases as we are asking about self-

486
reported illness and information may be inaccurate.

487
There is a possibility that we could see ongoing outbreaks in the study year. If this happens,

488
we might try to identify these where possible and may consider excluding households from 489 this study where all or most members have been exposed, including a definition of a co-490 primary case.

492
End of study

493
The study will be declared as ended when the database is closed to recruitment -after one

494
year or when the maximum number of households has been enrolled.

77
This project will estimate the amount of secondary spread that occurs in the home and 78 potentially identify asymptomatic infections which might have a role in transmission. Risk 79 factors and characteristics associated with secondary spread will be described including any 80 differences in transmission between Cryptosporidium species.

81
Methods and analysis

82
The study will prospectively identify cryptosporidiosis cases from North West England and

83
Wales over one year and invite them and their household to take part. Each household will 84 complete a questionnaire and each household member will be asked to provide a stool 85 sample. Clinical, demographic, and home variables will be described, and further analyses 86 undertaken to investigate associations with secondary spread in the home.

87
Cryptosporidium-positive stool samples, identified by immunofluorescence microscopy, will 88 be characterised using molecular methods to describe patterns of transmission. Data 89 collection is expected to take one year, beginning in September 2018.   1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59   157 hominis and C. parvum respectively) and found that contact with another person with 158 diarrhoea was a risk factor for infection with Cryptosporidium, and that exposure through 159 changing children's nappies was a specific risk factor for infection with C. hominis whether 160 the child was symptomatic or not. cases may subsequently arise following exposure to either, and outbreaks in close settings 182 such as the home or institutions may happen more frequently than is currently recognised.

183
Few studies exist which refute or confirm this, especially in industrialised countries.

185
The aim of this study is to estimate the amount of onward spread of Cryptosporidium that 186 happens in the home, and to describe associated factors and case characteristics. ( The identification of cases, and their subsequent recruitment, is cross-sectional, although the 237 study also involves retrospective data collection and some prospective sampling (Error!

239
Cases of cryptosporidiosis will be identified via the relevant surveillance system(s Case definitions Index case The first case from a household identified in the surveillance system (person reported to a PHE/PHW surveillance system(s) following detection of Cryptosporidium sp. in a faecal sample, with a specimen date in the study year)

473
Confounding (e.g. host factors such as age, co-morbidity) will be considered, where known, 474 using multivariable techniques. Also, we will, where possible, examine environmental level 475 exposures using stratification, e.g. those households/cases which are exposed to other 476 known sources or risk factors, such as those living on farms.

477
Data will be analysed using Stata v12.

LIMITATIONS AND BIASES
480 Some elements of the study design are retrospective in nature, as the index case must have 481 already been ill and been tested in order to be selected. As a result, some ascertainment 482 bias may lead to a skewed sample from which to choose the index cases. We do not expect 483 to capture the full profile of cases and households in the population that might have

484
Cryptosporidium due to differences in risk or vulnerability, severity, or health-seeking 485 behaviour [38]. We may get an over-representation of severe disease as these cases are 486 more likely to seek health care and be tested, and perhaps more likely to test positive.

487
We are only collecting one sample from each household member, and not re-sampling the 488 index case, for time and resource reasons. This may well lead to missing intermittent 489 shedding of oocysts, tertiary household infections and/or misclassifying recurring illness.

490
As Cryptosporidium is common in younger age groups, we expect a large proportion of the 491 participants to represent families with young children which may lead to over-representation 492 of these households. In addition, we might expect that having young children who were ill, or 493 being severely ill themselves, may incentivise cases to participate in the study, more than 494 adult, less severe cases.

495
Any likely over-or under-representation in the data collected will be considered when 496 assessing and describing results. Further unidentifiable limitations may include recall biases 497 around dates of onset or activities, and classification biases as we are asking about self-

498
reported illness and information may be inaccurate. There is a possibility that we could see ongoing outbreaks in the study year. If this happens,

500
we might try to identify these where possible and may consider excluding households from 501 this study where all or most members have been exposed, including a definition of a co-502 primary case.

504
End of study

505
The study will be declared as ended when the database is closed to recruitment -after one 506 year or when the maximum number of households has been enrolled.

507
Pilot arrangements

508
A pilot phase of one month is anticipated before data collection begins to assess and 509 evaluate processes. Pilot data will be included as study data if no major methodological 510 changes are proposed.

511
Ethics and dissemination

512
The study has been approved by the North West -Liverpool East NHS Research Ethics