Seeking help for peritoneal dialysis- associated peritonitis: Patients’ and families’ intentions and actions. A mixed methods study

AIMS
To examine patients' and families' help-seeking intentions and actions when suspecting peritoneal dialysis-associated peritonitis.


DESIGN
A sequential explanatory mixed methods design was used, comprising a questionnaire and semi-structured interviews.


METHODS
A questionnaire was designed, piloted and used with patients and family members (n=75) using peritoneal dialysis from six hospital sites in Wales and England. Questionnaire data were analysed using descriptive statistics. A purposive sample of questionnaire participants (n=30) then took part in telephone or face-to-face semi-structured interviews. Interview data were analysed thematically. Data were collected between September 2017 and August 2018. Ethical and governance approvals were obtained; the study was reported on national research portfolios.


RESULTS
The quantitative data highlighted differences between participants' knowledge of when they should seek help for suspected peritonitis and their actions when they subsequently experienced peritonitis. The interview data revealed the complexities involved with recognizing peritonitis, making the decision to seek help and accessing healthcare. Some participants struggled to recognize peritonitis when signs/symptoms started, leading to delays in deciding to seek help. Furthermore, some participants reported that they accessed help from renal or generic out-of-hours and were misadvised or misdiagnosed, delaying diagnosis and treatment. The data were integrated using conceptual analyses of help-seeking behaviour and access to healthcare, which informs understanding of the complexity of seeking help in this context.


CONCLUSIONS
This study revealed differences between participants' help-seeking intentions and actions. Using the conceptual analyses of help-seeking behaviour and access to healthcare informs understanding of the complexity of the help-seeking process in this context. To safely use a home therapy, it is imperative that individuals recognize signs/symptoms of peritonitis, seek help promptly and are appropriately supported when they access healthcare. Further work is needed to examine how these individual and system changes can be enacted.


| INTRODUC TI ON
Globally, kidney failure has a major impact on health, and the prevalence of this condition continues to increase (GBD Chronic Kidney Disease Collaboration, 2020). Kidney failure is irreversible but can be managed with peritoneal dialysis (PD), haemodialy-  (Worsey, 2019).
Peritonitis (infection of the peritoneum) is a serious problem for patients using PD. A recent multinational nominal group technique study with patients and carers reported that PD-related infection was their primary concern due to the potential serious consequences on health (Manera et al., 2019). Peritonitis is the cause of death for 4% of patients using PD and a contributing factor for 16%, and can lead to peritoneal membrane failure and, thus, withdrawal from PD (Li et al., 2016). The signs/symptoms of peritonitis include pyrexia, abdominal pain and cloudy dialysis effluent. A diagnosis of peritonitis is made when two of the following are present: 1. Clinical features of peritonitis, that is, abdominal pain and/or cloudy dialysis effluent; 2. Dialysis effluent white cell count >100/μL (after a dwell time >2 h), with >50% polymorphonuclear; 3. Positive dialysis effluent culture (Li et al., 2016).
However, patients with cloudy effluent should be presumed to have peritonitis and treated accordingly until a diagnosis is established (Li et al., 2016). Peritonitis can be caused by modifiable (skin or environmental contamination, catheter-related) and nonmodifiable (bowel or gynaecological flora, or bacteraemia) factors, and steps to prevent peritonitis are, therefore, vital (Worsey, 2019).

| BACKG ROU N D
A detailed review of the literature is reported in the published study protocol (Baillie et al., 2018). There is limited evidence considering the impact of peritonitis on the individual and their family. A small, dated number of quantitative studies reveal a relationship among peritonitis, depression, anxiety and quality of life (Bakewell et al., 2002;Juergensen et al., 1996Juergensen et al., , 1997Troidle et al., 2003). More recent qualitative studies show that peritonitis is an upsetting, painful and embarrassing experience for patients (Baillie & Lankshear, 2015b;Campbell et al., 2016). Several quantitative and qualitative studies have reported patients' and relatives' knowledge of peritonitis, and highlighted gaps related to peritonitis prevention, monitoring and identification (Baillie & Lankshear, 2015b;Campbell et al., 2016;Russo et al., 2006;Sayed et al., 2013). There is, however, a dearth of UK studies quantifying patients' and relatives' knowledge of peritonitis. International studies demonstrate that patients with increased knowledge about PD and peritonitis are less likely to develop the complication (Kazancioglu, Ozturk, Ekiz, et al.,

K E Y W O R D S
help seeking, infection, kidney failure, mixed methods, nursing, peritoneal dialysis, peritonitis, qualitative Impact • What problem did the study address? Peritonitis is a significant complication for patients using peritoneal dialysis, but a dearth of evidence exists about the helpseeking behaviour of this patient group.
• What were the main findings? The quantitative data demonstrated differences between participants' helpseeking intentions and actions when they suspected peritonitis. The qualitative data revealed difficulties in some participants recognizing the signs/symptoms of peritonitis, delays in deciding to seek help and subsequent difficulties in accessing appropriate healthcare.
Using the conceptual analyses of help-seeking behaviour and access to healthcare informs understanding of the complexity of seeking help in this context.
• Where and on whom will the research have an impact? For patients to safely use peritoneal dialysis at home they need to be able to recognize peritonitis and promptly seek help. Healthcare professionals need to be able to signpost patients with peritoneal dialysisassociated peritonitis to appropriate specialist services in a timely manner.
On suspicion of peritonitis, patients need to seek urgent assessment from the PD team (Salzer, 2018). Seeking help, however, is a complex phenomenon. A conceptual analysis of help-seeking behaviour in healthcare literature identifies it as a process in response to a problem that cannot be self-managed, and involves the need to interact with a third party (Cornally & McCarthy, 2011).
The antecedents for help seeking include: the individual recognizing and diagnosing a problem, making a decision to act and selecting a source of help (Cornally & McCarthy, 2011). Two qualitative studies with patients using peritoneal dialysis highlighted that individuals may struggle to recognize the signs/symptoms of peritonitis (Baillie & Lankshear, 2015a;Campbell et al., 2016). On identifying suspected peritonitis, the individual needs to access help from healthcare services. A systematic review highlighted a dearth of studies that have examined patients' help-seeking behaviours in response to suspected peritonitis (Griva et al., 2014) and a more recent literature search demonstrated no more recent evidence available in this field. Although not looking specifically at patients experiencing peritoneal dialysis, a recent UK qualitative study of patients with a complex condition demonstrated the challenges of accessing specialist secondary care in an emergency (Clarke et al., 2015). Levesque et al., (2013) conceptualize access as 'the opportunity to identify healthcare needs, to seek healthcare services, to reach, to obtain or use health care services and to actually have the need for services fulfilled' (p.8). Successful access of healthcare services is dependent on the ability of the individual/household/community/population and the accessibility of the provider/organization/institution/system (Levesque et al., 2013). Therefore, for patients to receive treatment for peritonitis, there are multiple processes that need to be followed.
This article explores the help-seeking intentions and actions of patients and their family members when they suspect peritoneal dialysis-associated peritonitis. These findings form part of a larger study that was designed to examine patients' and families' knowledge and experiences of PD-associated peritonitis, details of which can be found in the study protocol (Baillie et al., 2018).

| Aims
The aim of this article is to examine patients' and families' helpseeking intentions and actions when suspecting peritoneal dialysisassociated peritonitis.

| Design
A sequential explanatory mixed methods study (Cresswell & Creswell, 2018) was planned to meet the study aim and objectives. The questionnaire, administered first, enabled the collection of quantifiable information from a larger sample, while the semi-structured interviews generated rich data that added depth to the questionnaire data (Cresswell & Creswell, 2018;Kroll & Neri, 2009). Synthesizing the quantitative and qualitative data is a crucial stage of a mixed methods study (Kroll & Neri, 2009) and in a sequential explanatory mixed methods study should occur at the interpretation phase (Cresswell & Creswell, 2018). The results were, therefore, synthesized in the Discussion, drawing on relevant theory on help-seeking behaviour.

| Sample/participants
3.3.1 | Questionnaire recruitment and sample Participants were recruited from six National Health Service (NHS) organizations in Wales and England. A renal or research nurse at each site identified eligible participants, according to the inclusion criteria (Box 1), and provided them with an information pack. Potential participants returned a permission to contact form to the nurse if they were interested in participating; the nurse then returned these to the first author. The first author then telephoned the potential participant to explain the study, answer any questions and arranged a time to complete the telephone questionnaire. Relatives were informed about the study by the patient.

| Semi-structured interview recruitment and sample
A maximum variation purposive sample (Patton, 2015), aiming for variation in terms of age, time using PD, peritonitis diagnosis and site, was used to recruit those participants who completed the questionnaire, and agreed to a follow-up interview.

BOX 1 Inclusion criteria
Questionnaire: 1. Over 18 years old; 2. Able to give informed consent; 3. Using PD (continuous ambulatory peritoneal dialysis [CAPD], automated peritoneal dialysis [APD], or both); or used PD within one year of study start date, but now using haemodialysis or with a kidney transplant; 4. An adult (>18 years old) responsible for PD of a person meeting criteria 1 and 3; 5. Able to read and write in English.
Interviews: 1. A participant from Phase 1; 2. Able to speak in English; 3. Able to participate in a telephone or face-to-face interview.

| Questionnaire
A questionnaire was developed with reference to: previously used questionnaires Ozturk et al., 2009;Russo et al., 2006;Sayed et al., 2013); clinical guidelines (Li et al., 2016); renal textbooks (Levy et al., 2016;Main, 2014); patient literature (Oakley, 2016); Office for National Statistics guides (Office for National Statistics, 2015a, 2015b); UK Renal Registry report (MacNeill et al., 2016) and dialysis company literature. Furthermore, the questionnaire content was reviewed by renal doctors and nurses, researchers, patients, carers and a statistician. The questionnaire was piloted with the first 10 participants at the first study site to ensure the questions were clear (Sapsford, 2007). As only minor edits were required, these responses were included in the main study data collection. This process is outlined in more depth in the published protocol (Baillie et al., 2018) and a summary of the focus of questions is in Box 2 Multiple choice and dichotomous yes/no questions were used-a suitable approach for knowledge questionnaires (Rattray & Jones, 2005). However, some questions provided space for free-text comments, to ensure the questionnaire captured participants' knowledge or experience if different from the options provided.
The questionnaire was administered by the researcher over the telephone, between September 2017 and August 2018.

| Semi-structured interviews
Participants were interviewed using a semi-structured approach between March and July 2018 about their knowledge, understanding and experience of peritonitis. Interviews were via telephone or in person; these approaches have been successfully used with this population in previous studies (Baillie & Lankshear, 2015b;Campbell et al., 2016).
Patients and relatives were given the option of being interviewed together or separately, according to their preference. An interview guide (Box 3) was developed from the literature and incorporating emerging results from the questionnaire. For example, in the free-text comments of the questionnaire, participants highlighted difficulties in accessing help out-of-hours-this was explored in the interview guide. Ultimately, the interviews were iterative, for example, participants in the early interviews discussed strategies to prevent peritonitis when away from home, which was then discussed with later participants. Participants were recruited and interviewed until no new themes were identified (Guest et al., 2006). Interviews were digitally audio recorded and lasted an average of 31:03 min; interviews with participants who have experienced peritonitis lasted longer than interviews with participants who have not experienced peritonitis (mean: 38:28 min compared to 24:43 min). Audio recordings were transcribed verbatim by a professional transcriber and the transcriptions were checked by the first author.

| Ethical considerations
The study was undertaken in accordance with the UK Policy informed that their participation was voluntary and they could withdraw at any point without reprisal. Informed consent was taken prior to each questionnaire and interview. All data were anonymized and stored in line with legislation (United Kingdom, 1998, 2018. Interview participants were assigned a pseudonym alphabetically.

| Data analysis
This study generated two data sets, which were analysed separately, as per the requirements of a sequential explanatory mixed methods study (Cresswell & Creswell, 2018).

| Questionnaire data analysis
Data were downloaded from Online Surveys into Excel and inputted to a coding book created in SPSS 25 (IBM Corp., 2017). The data were analysed descriptively, presented in tabular form as numbers with percentages. The free-text data were extracted from Excel and analysed for key topics.

| Interview data analysis
Data analysis commenced during data generation, ensuring an iterative approach (Hammersley & Atkinson, 2007). The data were managed using NVivo 11 (QSR International Pty Ltd., 2015) and the data were analysed thematically (Wolcott, 1994), which involved generating a coding framework and coding the data (Description), considering the meaning of the data and identifying themes (Analysis) and finally interpreting these themes and considering them in relation to the wider literature (Interpretation).

| Interpretation
The conceptual analyses of Cornally and McCarthy (2011) and Levesque et al., (2013) were used to interpret the synthesized study findings, outlined in the Discussion.

| Rigour
To promote integrity and quality in this mixed methods study, appropriate validation strategies were applied for each phase of the study (Cresswell & Creswell, 2018;Giddings & Grant, 2009). These are summarized in Table 1. Furthermore, Cresswell and Creswell (2018) assert the importance of selecting the sample for the qualitative phase from the questionnaire participants. This was undertaken in the current study.

| FINDING S
The findings for each phase of the study are presented in turn. Table 2.

Demographic and clinical information about participants is in
The questionnaire data revealed differences between patients' helpseeking intentions and actions when peritonitis was suspected. The interview data enabled these differences to be further explored.

| Questionnaire findings
In total, 75 participants (patients n = 61, relatives n = 14) completed the telephone questionnaire. The questionnaire results are

Quality marker How the marker was achieved
Phase 1: Questionnaire

Content validity
Questionnaire items generated with input from experts in the field (Rattray & Jones, 2005).

Pilot
Questionnaire piloted in the first site, identifying questions that needed slight amendment (Rattray & Jones, 2005).

Transferability
Demographic information about each participant. Asked sites to provide contextual information about their PD service (not all sites completed the information).

Dependability
Completion of research journal, documenting the audit trail (Koch, 1994) Confirmability Reflective journal maintained to promote reflexivity.

TA B L E 1
Promoting rigour in the study.
summarized in Table 3. Participants were asked to identify the signs and symptoms of peritonitis and the actions they would take if they suspected peritonitis. The participants who experienced peritonitis (n = 28, 37.3%) were asked to list the signs/symptoms that alerted them to peritonitis and the actions they took on suspicion of peritonitis.

TA B L E 2 Participant demographic and clinical information.
Questionnaire (20%) patients stated that they 'waited to see how they felt', which is higher than the 1 (1.6%) patient who reported that they would do this if they suspected peritonitis. Interestingly, in the free-text comments, six participants reported that they waited to contact the clinical team, which included waiting for their next exchange (n = 1), waiting until morning (n = 2) and waiting 1 day (n = 3). Therefore, multiple participants waited to seek help when they first suspected peritonitis. Furthermore, two participants did not experience symptoms and their peritonitis was identified via routine blood tests. One participant reported that they contacted the clinical team but were told (incorrectly) it was not peritonitis and to call an ambulance. Four (20%) patients reported that they saved their drained bag of dialysis effluent, which is consistent with the 13 (21.3%) who had intended that they would do this, but again not doing so may hinder the diagnosis process. All relatives previously reported that they would contact the clinical team if they suspected their relative was experiencing peritonitis; almost all (n = 7, 87.5%) relatives stated that they contacted the clinical team when this happened. The one participant who selected 'no' to 'contacted the clinical team' explained that her husband was an inpatient on a non-renal ward when she identified his effluent was cloudy. Therefore, in reality, all relatives raised their concerns about peritonitis with healthcare professionals. No relatives reported that they waited to see how their loved one felt, which is consistent with their intended actions reported above.

| Interview findings
Thirty participants who completed the questionnaire also took part in a semi-structured interview, including 23 patients and 7 relatives.
Fifteen participants had experience of peritonitis, including 11 patients and 4 relatives. The following themes and their sub-themes are presented: •

| Confusion
Crucially, some participants reported confusion about recognizing signs/symptoms of peritonitis. In some cases, the patient or relative reported confusion, but in other cases when individuals sought help, they were misinformed. Iris developed peritonitis shortly after her abdominal catheter (dialysis access) was inserted into her abdomen, and attributed abdominal pain to the procedure: 'First of all I thought "Oh, I am really sore internally", and I know that, they'd then just got the tube in, and I thought maybe it's, so that was that. Then I thought: "No I will mention it, it's very, very uncomfortable now internally." After a few days… I thought "I really must tell them, it's very uncomfortable"' (Iris, patient) In the section above, Kieran described severe abdominal pain that alerted him to peritonitis, but he had identified a cloudy bag earlier that evening. He explained that he could not feel certain the bag was turbid and, therefore, waited for the next dialysis exchange: While Catherine felt able to wait until the next morning, Kieran became very unwell and went to the hospital in the middle of the night.
Kieran described his concern that he would 'waste' the hospital's time and resources, or cause disruption to his wife and son unnecessarily.
However, Kieran did acknowledge that he would now always ring for help if he had concerns, demonstrating an increasing trust in his ability to self-assess and self-manage his dialysis: "it was late at night, it looked slightly cloudy. I don't really know if it is going to be a problem or not. Maybe the next bag will tell me for sure, so I will wait a bit and after, Therefore, many participants delayed seeking help, from a few hours to 1 week. While some participants described learning from this delay and vowed to report their concerns more promptly in the future, other participants reported that they continued to wait to ask for help, including Bridget who had experienced multiple episodes of peritonitis, and continued to wait for several days to get help: 'I think in the first I could feel that probably something wasn't quite right because my tummy used to feel sore. Um, but if I left it more than a couple of days then it did get quite painful.' (Bridget, patient)

| Challenges in accessing help
All 30 participants, from six NHS services, knew they needed to contact the PD nursing team if they suspected peritonitis. Out-of-hours, generally between 5pm and 9am and over the weekend, participants were aware they could contact the renal ward for advice, rather than using emergency services: 'if they're [PD team] not there after 5, they work till 5, you've got [renal ward] to go to. Um, I've got their ward number as well, or go straight to them as well, so the ward for kidney patients, so up there as well. So, there's lots of places to go to, rather than go to A&E.' (Iris, patient) However, multiple participants reported challenges in accessing healthcare from the renal ward out-of-hours. This may be due to unfamiliarity with the renal ward and concerns about the specialist PD knowledge of general renal healthcare professionals. The latter is problematic as PD is a specialist treatment, but patients are advised to contact a non-PD service out-of-hours. Kieran highlighted challenges in attending hospital out-of-hours when he was feeling very unwell. Kieran's experience demonstrates difficulties patients may face in getting to hospital (access to a car may be needed, with someone else to drive it) and finding the correct ward if they are unwell and potentially alone: Derek's wife Doreen reported being misadvised when she telephoned the renal ward out-of-hours to report Derek's severe abdominal pain, leading to a delay in checking the PD effluent. This resulted in a long wait overnight for an ambulance before Doreen was able to speak to a PD nurse and Derek was taken straight to the PD unit by an ambulance. Following this 12-hour delay, Derek was hospitalized for 2 weeks with peritonitis: 'so around about 9 o'clock, as soon as I could really, half past 8 to 9, something like that, I rang the ward again and… I then spoke to one of the PD nurses who had come on duty. Um, and she said: "Right, the first thing, um, have you done a bag?" And I said: "Well, no" and she Fortunately, Derek was taken by the paramedics straight to the PD unit, rather than the emergency department, which enabled Derek to start antibiotics more promptly than Adam, whose wife Audrey also faced challenges in accessing healthcare. There are similarities between the experiences of the two couples, in terms of long ambulance delays, and poor advice from the renal wards at the point Therefore, when some participants eventually sought help from out-of-hours renal services, they experienced difficulties in attending the hospital, finding the ward and receiving appropriate care from staff who were not familiar with peritoneal dialysis.

| DISCUSS ION
This study revealed differences between participants' help-seeking intentions and actions on suspicion of peritonitis. We believe that this is the first study to examine help-seeking intentions in this population. To safely use a home therapy, it is imperative that individuals recognize when to seek help and can promptly access healthcare when required. This discussion will draw on the conceptual analyses of Cornally and McCarthy (2011) and Levesque et al., (2013), and also renal and wider help-seeking behaviour literature, in the absence of studies examining help-seeking behaviour in response to PD-associated peritonitis (Griva et al., 2014). (2011) (Russo et al., 2006;Sayed et al., 2013). However, participants in the current study were not always able to recognize peritonitis signs/symptoms when they occurred. An Australian qualitative study similarly showed that patients may struggle to interpret their peritonitis symptoms (Campbell et al., 2016), a finding earlier identified in a UK ethnographic study (Baillie & Lankshear, 2015a). Interestingly, participants in the current study also reported that PD healthcare professionals at times struggled to recognize whether a PD effluent bag was turbid, highlighting the potential subjectivity of a 'cloudy bag'. This is concerning as turbid effluent is a key diagnostic indicator for peritonitis (Li et al., 2016). However, a recent observation study of 247 patients identified that older people (>65 years) were less likely to present with pyrexia or cloudy PD effluent (Htay et al., 2019), the latter they attribute to difficulties in recognizing a cloudy bag. The authors concluded that healthcare providers must have a high index of suspicion for diagnosing peritonitis in older people. Danguilan et al., (2013) reported that patients in their education programme had low knowledge of the signs/symptoms of kidney failure and this resulted in delays in help seeking. They, therefore, highlighted the need for public education about the signs/symptoms of CKD to avoid delays in referral to healthcare professionals. Therefore, while it is clear that patients and families need additional support to recognize the signs/ symptoms of peritonitis, Cornally and McCarthy (2011) assert that interventions aimed at improving knowledge only address one aspect of the help-seeking process-problem recognition-and the factors affecting the decision to seek help and the source of help also need addressing to improve help-seeking behaviour.

Cornally and McCarthy
The next phase of Cornally and McCarthy's (2011) help-seeking process involves making a decision to act. In the current study, it was evident from the findings of both phases of the study that respondents were aware that if they suspected peritonitis, they should contact their clinical team immediately. Peritonitis can be fatal and requires urgent attention (Salzer, 2018); therefore, in patients presenting with either a cloudy bag or abdominal pain (in the absence of the other) the differential diagnosis should be peritonitis, until this is confirmed or excluded (Li et al., 2016). However, when participants in the current study recognized that they may have peritonitis, multiple individuals in both phases of the study waited to act. Often this was due to uncertainty about whether the sign or symptom recognized was peritonitis, and participants did not want to 'waste' NHS resources unnecessarily. A grounded theory of help seeking in relation to urinary incontinence in primary care highlighted that patients may experience symptoms, reappraise unresolved symptoms and then seek help (Shaw et al., 2008). However, for patients with suspected peritonitis, this is an emergency that needs prompt treatment. Clarke et al., (2015) highlighted patients delayed seeking help for an average of 11 hours after symptom onset for neutropenic sepsis, partly due to concern about wasting NHS resources and hope that symptoms would resolve. Similar to patients using peritoneal dialysis, patients at risk of neutropenic sepsis are told when to seek help, but again the study by Clarke et al., (2015) demonstrates this process is not simplistic. Therefore, an intervention is needed to support patients and families to make the decision to act on first recognition of peritonitis signs/symptoms; this is vital to promote patient safety.
Once participants have recognized potential peritonitis and decided to act, they were required to select a source of help-the next phase of Cornally and McCarthy's (2011)  Family members were required to drive their loved one, participants faced long journeys to specialist centres and some patients required an ambulance due to the severity of the peritonitis. Campbell et al., (2016) similarly identified the important role played by family members in supporting patients once peritonitis was suspected, including driving them to hospital. There were also challenges in this when patients were advised to go through the emergency unit, thus, limiting their access to the specialist service they needed. Participants in this study reported being very unwell trying to get from the car park to the ward out-of-hours, this was also showed by Clarke et al., (2015). The affordability of accessing healthcare for peritonitis will vary according to the country and the health costs and social support available in that country. Participants did not raise affordability in this UK-based study, although many participants were required to travel long distances to their peritoneal dialysis centre, which does incur financial costs. An Australian study further highlighted the financial impact of seeking help for peritonitis, including taking time off work and cost of frequent hospital appointments for peritonitis treatment (Campbell et al., 2016). Furthermore, there are potentially significant cost implications for the NHS if delays in diagnosing peritonitis occur. It is essential that peritonitis is promptly and appropriately managed (Salzer, 2018). In the Levesque et al., (2013) model, appropriateness refers to the service provided and the quality of that service. Participants in the current study reported variations in the care they received, particularly when accessing help out-ofhours. Similarly, Clarke et al., (2015) highlighted that non-specialist healthcare professionals' lack of knowledge and logistical problems such as unanswered telephones led to delays in urgent treatment for patients with neutropenic sepsis. Overall, participants in the current study encountered problems when accessing help out-of-hours and this urgently needs addressing in clinical practice to ensure patients and families can access support in a timely way.

| Limitations
The size of the non-probability sample in the quantitative phase of the study is a limitation and affects the generalizability of the results; steps were taken to increase the sample, including changing the approach to recruitment and the inclusion of additional sites. Utilizing an online questionnaire may have increased the sample size and reduced potential social-acceptability bias. However, the questionnaire data set was complete, and the telephone questionnaire was administered by the first author, ensuring consistency. The mixed methods approach allowed for triangulation of methods and validation of data as well as enabling a richer picture and greater insights into quantitative findings. Synthesizing the quantitative and qualitative data,

| CON CLUS IONS
This study has identified the help-seeking intentions and actions of families when they suspect peritonitis-and the disparities between these. Additionally, the qualitative work demonstrated the challenges participants have in recognizing peritonitis, deciding to seek help and accessing appropriate and timely care.
Using the conceptual analyses of help-seeking behaviour and access to healthcare informs understanding of the complexity of seeking help in this context. If patients and families are to safely self-manage complex medical treatments at home, they must be supported to recognize complications and know when to ask for help. Furthermore, they must be able to access prompt medical treatment from their tertiary NHS organization on suspicion of peritonitis. Further research is now needed to ascertain the best ways to support patients and families to recognize complications and decide to seek help in a timely manner. Crucially, change is required in clinical practice to ensure patients receiving specialist treatments can access appropriate care out-of-hours.

CO N FLI C T O F I NTE R E S T
No conflict of interest has been declared by the authors.