Introduction

Small Intestinal Bacterial Overgrowth (SIBO) is characterized by excessive and/or abnormal bacterial growth in the small intestine, defined by populations exceeding ≥ 103 bacteria (i.e. colony-forming units [CFU]) per mL of proximal jejunal aspiration. This most commonly results from disruption in the secretion of gastric acid and/or small intestinal dysmotility, with immunological conditions and anatomical obstructions representing significant risk factors [1,2,3].

SIBO presents with a wide variety of signs and symptoms and has a significant negative impact on patients’ Health Related Quality of Life (HRQoL). SIBO patients experience a range of intestinal and extraintestinal symptomatology including diarrhea, nausea, bloating, abdominal pain, vomiting, and flatulence; the condition can further result in malabsorption, malnutrition, weight loss, and other sequelae [4]. While the general prevalence of SIBO is unknown, prevalence in elderly populations may be as high as 15% and even higher in elderly patients with other risk factors (e.g. diminished acid production) [4]. SIBO is thought to be substantially underdiagnosed due to symptom heterogeneity, frequent co-occurrence with other gastrointestinal (GI) disorders (e.g. irritable bowel syndrome [IBS], diarrhea, and constipation), and challenges related to diagnostic procedure [4,5,6].

Current management strategies include identifying and ameliorating underlying conditions, addressing nutritional deficiencies, and using oral antibiotics [7]. Symptomatic SIBO is typically treated with antibiotics, including rifaximin, designed to reduce bacterial overload. The problem remains that antimicrobial therapies continue to be sub-optimal in terms of efficacy, and safety, with the promise of better results using formulations that combine immediate release with sustained release drug delivery systems [8].

The development of novel treatments has been hampered by protocols that differ widely between studies, with differences across study populations, diagnostic tests and interpretation, and dosing and duration of antibiotic therapy [8]. Adequate and well-controlled studies are needed to substantiate previously reported findings about improved formulations and to establish an optimal treatment regimen [9]. Further, well-defined and reliable endpoint measures are needed to conduct pivotal work.

Currently, there is no disease-specific patient-reported outcomes (PRO) measure for SIBO patients that assesses symptom burden, is fit-for-purpose, and has been developed in accordance with United States (US) Food and Drug Administration (FDA) guidance [9, 10]. The aim of this study was to address this unmet need, constructing a content valid PRO measure suitable for assessing symptom severity in clinical trials with SIBO patients.

In this study, we detail qualitative research to: 1) understand the symptoms that are most bothersome and important to SIBO patients, and 2) develop a PRO measure suitable for assessing symptom severity in clinical trials with SIBO patients. The broader goal of this study is to establish content validity for the new, condition-specific measure called the SIBO Symptom Measure (SSM).

Methods

Patient inclusion and recruitment

Patients (≥ 18 years) for all PRO development stages (Stage 1–3 described below) were recruited through four clinical sites across the US. Sites identified patients by reviewing medical records and patient databases and contacted candidates using IRB-approved recruitment materials. Candidates were screened by clinical site personnel according to the inclusion/exclusion criteria (Table 1). A positive lactulose hydrogen breath test (LHBT) result was required to confirm diagnosis. Following successful screening, patients attended the clinic to sign an informed consent form outlining the study purpose/goals. No relationship was established with patients prior to the interviews. Subsequently, patients were scheduled for their interviews and compensated for their participation. Patient interviews in all stages were audio-recorded, transcribed verbatim, and anonymised prior to data analysis. Interviewers had 5 + years of qualitative interviewing experience and interviewer trainings (on interview guide/study procedures) were conducted prior to interviews.

Table 1 Inclusion and exclusion criteria
Full size table

Study design

The study consisted of three primary stages. Targeted literature review (TLR), clinician interviews, and one-on-one patient interviews in Stage 1 identified SIBO symptoms and concepts important to patients for preliminary inclusion in the SSM. In Stage 2, additional patients were interviewed to learn about their SIBO experience and test the draft SSM. In Stage 3 the draft SSM was refined based on patient feedback and further tested for its content. Specifically, patient feedback on the draft SSM systematically evaluated each item in terms of ease of completion, item and instruction interpretation, and appropriateness of response options (both verbal rating scale and numeric rating scale were tested); patients were also asked if any key symptoms were missing. The main steps involved in the development of the SSM are presented in Fig. 1.

Fig. 1
figure 1

Study flow. This study comprised of 3 stages which consisted of various interview waves, feedback from clinicians, and guidance from the FDA. This work resulted in a content valid measure, the Small Intestinal Bacterial Overgrowth Symptom Measure (SSM). CE concept elicitation, CI cognitive interviews, FDA US Food and Drug Administration, SSM Small Intestinal Bacterial Overgrowth Symptom Measure

Full size image

Stage 1—literature review, clinician interviews, and initial concept elicitation interviews

The initial study phase comprised of a TLR, clinician interviews, and brief concept elicitation (CE) patient interviews to explore salient concepts in SIBO for draft SSM development.

Two TLR searches (using combination of controlled and free-text terms [11]) were conducted in Medline, Embase, and PsycInfo to identify English-language publications from January 2014 to January 2019 (conducted in January 2019) comprising of adult SIBO patients (≥ 18 years) that reported on symptoms, impacts, functioning and/or HRQoL related to disease burden (Table 2), and included a PRO measure utilised for SIBO patients (Table 3). Case reports, letters, and editorials were excluded (Table 4). Based on clinician inputs (who are also co-authors), the definition of SIBO has evolved over time and as such older literature may not capture the current thinking about the disease. For example, the first North American consensus on how to diagnosis SIBO utilizing methane-based breath testing was published by Rezaie et al. in May 2017 [2]. Further, the American College of Gastroenterology SIBO Clinical Guidelines by Pimentel M, et al. were published in February 2020 which shapes the current thinking of the disease [9]. Hence, the TLR was designed to identify the most recent 5 years of data i.e. January 2014 to January 2019. These results were used to inform the clinician interview guide and provide insights into potential symptoms and impacts to include in the first draft development of the SSM. A follow up search in May 2022 was conducted to ensure most updated and relevant literature published from January 2019 to May 2022 was captured.

Table 2 Literature review search strategy to identify qualitative literature in SIBO
Full size table
Table 3 Literature review search to identify patient reported outcomes in SIBO
Full size table
Table 4 Eligibility criteria for inclusion in targeted literature review
Full size table

In-person, 60-min CE interviews were conducted with 8 patients and 3 clinicians, to gather an initial understanding of SIBO disease experience. The interview guide for both patients and clinicians included questions on experiences/knowledge of SIBO, key signs and symptoms, treatment experience and impacts on HRQoL and functioning. The data from these initial CE interviews, along with the TLR and clinician interviews, was used to develop the first version of the SSM to be tested in cognitive interviews (CI) conducted in stages 2 and 3. Further, findings from the TLR and clinical expert interviews informed the development of semi-structured patient interview guides for the CE interviews in stage 2.

Stage 2—combined concept elicitation and cognitive interviews

In stage 2, patient interviews were virtual, lasted 90-min, and included an abbreviated CE portion and a CI portion on the initial draft SSM. The purpose of the additional semi-structured CE interviews was to strengthen rationale for inclusion of each symptom in the SSM, as well as ensure comprehensiveness (i.e. saturation) of the items. Saturation allows evaluation of thematic saturation and is defined as the point at which no new concepts emerge [12]. CI interviews in this stage were conducted iteratively in three “waves” (of the total 4 interview waves) to allow for refinement of the draft SSM between interview waves. Following each wave, transcripts were qualitatively analysed; and each SSM item evaluated for clarity, applicability of response options, patient interpretation, and relevance. These analyses informed any necessary revisions to the SSM instructions, format, and items between waves, with changes tested to achieve a content-valid version of the SSM.

After stage 2, FDA feedback was sought on the SSM from the Clinical Outcomes Assessment Committee and the Office of Immunology and Inflammation, Division of Gastroenterology. The FDA provided their feedback as a Written Responses Only.

Stage 3—exclusively cognitive interviews

Stage 3 was conducted following FDA feedback on the SSM. Interviews in stage 3 (wave 4) were virtual, lasted 60-min, and were designed to confirm the content validity of the SSM in terms of patient understanding, ease of completion, and the relevance of the items/response options to the patients.

Data analysis

Interviews were analysed using an iterative, thematic analysis approach based broadly in the grounded theory approach [13]. This is a rigorous and transparent qualitative methodology which allows to achieve conceptual saturation through an iterative process of data collection and analysis. Analysis was conducted using the MaxQDA 2020 (VERBI Software, 2020) software. The software allowed the study team to assign labels or codes to sections of the transcripts (quotations) [13, 14]. As new data was collected, the study team added, defined, and refined the list of codes to develop a structured and comprehensive codebook. Within this codebook, codes were grouped into themes, defined by the content of the coded quotation [13, 14]. This approach was well-suited to capturing symptoms of direct importance to patients and sufficiently flexible to meet the validation objectives of CE [15]. This allowed the study team to confirm with study patients that the symptoms emerging from the interviews were meaningful, relevant, and accurately defined.

At an early stage of the analysis, a sub-section of the data (~ 20% of the transcripts) were double coded by a second analyst to increase the reliability and validity of the qualitative analysis. The coding was reviewed in a meeting with a third member of the study team, with the codebook revised accordingly. This double-coding approach built validity and reliability into the coding process and facilitated the assessment of saturation [12].

Prominent symptoms that emerged from the analysis were tracked using a saturation matrix to document conceptual coverage. Moreover, examination of these symptoms sought to provide an understanding of condition-specific symptoms most prevalent in SIBO patients. Descriptive statistics was used to analyse the collected sociodemographic and clinical data.

Results

Stage 1

Literature review

The TLR involved two searches (conducted in January 2019 and repeated in May 2022), one identifying qualitative research (search 1) and one identifying PROs measuring SIBO symptoms (search 2). Search 1 resulted in 66 records but screening did not identify any qualitative research studies in SIBO patients. However, SIBO has been proposed to be common in IBS patients with reported prevalence ranging from 10 to 70% [16]. As symptoms related to SIBO are somewhat similar to IBS symptoms, a second search was conducted to identify qualitative studies in IBS, yielding 69 references of which 5 articles were selected for data extraction (Supplementary Fig. 1). Symptoms reported included bowel symptoms, abdominal pain, abdominal discomfort, gas related symptoms, and other GI symptoms such as nausea, and non-GI symptoms such as fatigue.

Search 2 yielded 319 records after duplicates were removed based on the search criteria in Table 3. Upon review of the abstracts, 206 records were excluded. The remaining 113 records were selected for full text review, and 100 were selected for critical review (Supplementary Fig. 2). This literature review identified commonly used PROs such as the Bristol Stool Form Scale (BSFS) and the Patient Assessment of Upper Gastrointestinal Disorders-Symptom Severity Index; no SIBO-specific PROs were identified.

Clinician interviews

Three gastroenterologists with extensive clinical experience in diagnosing and managing SIBO patients were interviewed. All three gastroenterologists reported that bloating and distention are the most frequently reported and most bothersome symptoms (to patients) and the most important SIBO symptoms to treat. The gastroenterologists defined distension as the physical manifestation (i.e. enlarged abdomen) of excess gas, and bloating as the sensation of gassiness. They viewed bloating and distension as distinct concepts that may or may not co-occur when patients are describing their SIBO experience. They further noted that how patients report their symptoms depends on their baseline perception of “normal” as it relates to bloating and distension, thus making it difficult to think of a standardized way to describe these symptoms. Other symptoms reported by the gastroenterologists included diarrhea, constipation, dyspepsia, abdominal pain, discomfort, weight fluctuations, reflux, brain fog, fatigue, joint pain, and rectal symptoms.

Concept elicitation interviews (N = 8)

CE interviews were conducted both before the development of the draft SSM (Stage 1), and during refinement of the SSM (Stage 2). The first stage of CE interviews lasted 60 min and included 8 patients (mean age 46.2 years [range: 25–61 years], 7 females, 6 blacks; Table 5). All 8 patients spontaneously reported bloating as a key symptom (Table 6). Additional key symptoms that were reported by at least half of the sample and were identified as the initial core set of symptoms in the SSM included abdominal distention, gassiness, abdominal pain, appetite changes, fatigue, constipation, weight changes, and abdominal discomfort. Within these 8 patients, the following symptoms were reported as most bothersome: bloating (N = 4), distension (N = 2), abdominal pain (N = 2), diarrhea (N = 1), and fatigue (N = 1). Patients generally described their symptoms as varying throughout the day, and day-by-day. For example, 7 patients reported that bloating occurs daily or is constant and fluctuates throughout the day.

Table 5 Sample Demographics
Full size table
Table 6 Saturation grid (N = 23)
Full size table

Stage 2

Combined concept elicitation and cognitive interviews (N = 15)

CE interviews in Stage 2 were abbreviated due to inclusion of CI discussed further below and were conducted with 15 additional patients. For this stage, the mean age was 54.5 years (range 34–77 years) with 8 males and majority of white patients (N = 10) (Table 5). Combining stage 1 and stage 2 data, our sample of CE interviews includes a total of 23 SIBO patients. There were 4 new symptom concepts identified in the second stage: heartburn/acid reflux, joint pain, concentration, and shortness of breath (Table 6). However, each of these symptoms was reported by 1–3 patients and did not appear prevalent among the patient cohort as whole, nor were any of them clearly attributed to SIBO specifically. Table 6 shows the complete saturation grid from CE interviews, indicating that concept saturation was reached in terms of condition-specific symptoms by the 17th interview. Further, Table 7 shows descriptors patients used to define symptoms.

Table 7 Summary of patient-reported symptom descriptions, symptom definition, and draft item wording
Full size table

As previously noted, a total of 4 waves, each wave representing a modified SSM, were conducted within Stages 2 and 3 (three waves in Stage 2 and one wave in Stage 3). Conducting the CIs in waves allowed for the refinement of the measure as more patient data were collected. The evolution of the SSM is described below and as a condensed item tracking matrix (Supplementary Table 1).

Wave 1 represents the original version of the SSM before any changes were made. The SSM version in this wave of interviews included 9 items: bloating, abdominal distention, abdominal discomfort, abdominal pain, flatulence, fatigue, number of bowel movements, diarrhea, and constipation. Most patients (N = 14) in this stage found the SSM easy to complete and relevant to their experience with SIBO. There were some mixed definitions for the abdominal distension item, which was further tested as is in wave 2.

For wave 2 interviews, a nausea item was added due to the emergence of this symptom from wave 1 (N = 2) during the CE portion; for a total of 10 items on the SSM. Following this wave, a follow up consultation was conducted with the clinical experts (i.e. gastroenterologists) to gain feedback on the measure.

Following patient feedback from waves 1 and 2 and the clinicians’ inputs, wave 3 included the addition of a belching, appetite loss, and meal skipping item, for a total of 13 items. One clinician suggested adding the item on meal skipping, not intended for inclusion in the total SSM score but to examine in further psychometric work. In addition, the item on bowel movements (not intended to be included in the final score) serving as a gating item for subsequent diarrhea item, was included based on qualitative interviews and clinician inputs.

Most patients (N = 5) noted the questionnaire was relevant and accurately outlines how patients experience SIBO. The remaining patients (N = 3) noted the questionnaire was precise, simple, and relevant. Most patients (N = 14) were able to define the symptom of bloating, however, to help distinguish from the distension item, the wording was changed to “feeling of bloating”. Although there was some conceptual overlap between abdominal distension and bloating, we found enough support from both clinician and patient feedback to retain the two as separate items, until further psychometric evaluation. To help patients distinguish between these two symptoms, a definition of abdominal distention was added in this wave. All patients in wave 1 and 2 understood flatulence, however it was not clear how the severity of flatulence was determined. In wave 3, the flatulence item was revised so that the lower end of the scale represented “normal” flatulence.

Stage 3

Cognitive interviewing (N = 12)

Following FDA feedback, minor edits were made to the SSM, namely the flatulence and appetite loss items were modified to ask about worst levels as opposed to severity. For ease of understanding, the fatigue item was reworded to physical tiredness. Stage 3 comprised of the last SSM version (wave 4 of interviews), tested with 12 additional patients (mean age 37.5 years [range: 19–56 years], 6 females, and majority (N = 10) white) (Table 5). One patient in this wave expressed intellectual difficulties with understanding and interpreting the measure, most likely because of illiteracy. Therefore, 13 interviews were conducted, and one patient’s transcript was excluded from CI analysis. For defining diarrhea, the SSM references the BSFS stool consistency Types 6 and 7 [17, 18]. In wave 4, the diarrhea item was simplified, to include two relevant forms (types 6 and 7) of the BSFS, and the appetite item was re-worded to appetite loss rather than appetite, for clarity. Overall, patients in this wave found the questionnaire to be a suitable instrument to evaluate SIBO symptoms. Based on results from wave 4, no additional items were added and no additional changes were made.

Final SSM format for forthcoming psychometric validation

Version 1.0 of the SSM emerging from this process is an 11-item (13 items including meal skipping and number of bowel movements which are not included in total score) PRO instrument developed in accordance with the FDA guidance for developing new, content valid measures [10]. The novel SSM assesses severity of the key symptoms associated with SIBO. The SSM, which measures overall symptom severity as a single construct, assesses the severity of bloating, abdominal distention, abdominal discomfort, abdominal pain, flatulence, physical tiredness, nausea, diarrhea, constipation, appetite loss, and belching. With the exception of diarrhea, which is assessed as a daily frequency, response options range from 0 to 10, where 0 is the absence of the symptom and 10 is the worst possible experience of the symptom. Following CI, patients shared their SIBO experience changes on a daily basis, and that a daily recall period was appropriate. Using both CE and CI techniques, a content valid measure was developed. Psychometric testing to establish additional measurement properties in future studies will confirm reliability, construct validity, scoring metrics, and the ability to detect change for the SSM, allowing for final revisions to content that optimize measure performance for clinical trials use.

Discussion

The development of the SSM was based upon three types of evidence: review of published peer-reviewed literature, interviews with clinical experts, and CE/CI interviews with 35 SIBO patients. An in-depth understanding emerged of the important symptoms associated with SIBO across all three sources, providing both a fine-textured and comprehensive overview of the patients’ experience with SIBO. For example, bloating emerged as one of the most bothersome symptoms and was the most frequently reported symptom across the patient sample, which is consistent with the literature [6]. Abdominal pain was another important and prevalent symptom from the patient perspective, which is also mirrored in the literature [6]. We found some patients in our sample associated discomfort with bloating, as well as gassiness with these two symptoms.

As can often be the case in qualitative exploration of GI distress symptoms, important items may create problems for unidimensional measurement. For example, SIBO patients shared the importance of both diarrhea and constipation in their illness experience, which could provide more noise since both (or in some cases even one) of these items contribute to total scoring in use of the SSM as an endpoint measure.

At this point, projections about the optimal scoring of the SSM and its performance as a symptom measure remain empirical questions. While qualitative research ensures content that matters to patients, it often embeds ambiguities in measures requiring quantitative resolution. Items may be used in ways that were not anticipated despite careful attention to patient understandings and preferences for framing. Ultimately, future psychometric evaluations will allow for determinations about item performance. It may prove, for example, that retaining the discomfort item proves to be redundant with pain and/or bloating or that it may be useful to leave reports of constipation and/or diarrhea outside the calculation of the total symptoms score.

Based on patients’ interviews, nearly all items ask the patient to rate symptom severity at its worst level during the past 24 h (with the exception of the diarrhea item). We noted that in exploring the concept of diarrhea severity during the qualitative interviews, patients discussed stool consistency and urgency, in addition to frequency, which they felt was the best way to measure diarrhea severity; CI confirmed this item as relevant and easy to answer from the patient perspective. And as noted above, it remains an open question if this item, or any others will ultimately misfit the measure, prove internally inconsistent, or otherwise be used by respondents in ways that provide less information to symptom scoring than was anticipated.

Qualitative research may also provide a broader range of concepts having limited utility for reliable measurement. In this study, we found other signs and symptoms, such as weight loss, brain fog, vomiting, and heartburn/acid reflux, that we chose to exclude from the SSM after careful consideration. While these appeared not to be wholly idiosyncratic reports in some cases, their infrequent elicitations and failure to reappear in later interviews, provided for their exclusion.

Strengths and limitations

Consistent with the literature [19], our sample comprised of majority middle aged to elderly patients. Also, the SIBO clinical diagnosis was confirmed with LHBT testing, rather than relying on self-diagnosis. Recruitment through multiple clinical sites helped achieve greater diversity in our sample and did not limit us to one geographic area in the US. Our semi-structured interview guide was designed to elicit not only which symptoms patients experienced, but how symptoms fluctuated over time and the relative importance of each symptom in impacting the patient’s HRQoL. This yielded a rich and informative qualitative dataset to better understand how symptoms should be measured in this population. Lastly, the FDA feedback was sought and incorporated, which is advantageous for measure development.

Like most qualitative research, the sample size (n = 35 across all stages), while satisfactory [20] to establish content validity, was relatively small. Our sample included patients based only in the US, and thus reflects limited geographic and cultural diversity. Though no sex predilection for SIBO exists, our sample consisted of mostly college-educated, female patients, which may not be representative of the whole SIBO population. Further, this study did not include patient and public involvement (PPI) as part of the research process, which may be seen as a limitation. External validity always remains a concern in qualitative work, and despite careful attention to concept saturation, there may be elements of patient experience that remain uncaptured.

Finally, the measure has not yet been tested as a daily diary, and therefore the variation in scores between days, across a week, or through longer time intervals, will be tested in future psychometric work. As noted, reproducibility and other measurement properties will also be investigated in future psychometric work with the instrument.

Conclusions

Patient perspective is increasingly viewed as essential for evaluating treatment efficacy and making treatment decisions. Well defined and reliable PRO instruments are needed to enable physicians and researchers to evaluate the effects of treatments for SIBO. This study reflects a robust approach in developing a new PRO measuring symptom severity in SIBO that follows FDA guidance, has a short (24 h) recall period, and reflects patient concerns and is clearly understood by SIBO patients.

In summary, we conclude that the CE and CI data demonstrate content validity of the latest version of the SSM for measuring symptom severity among SIBO patients. Further psychometric testing will be needed to confirm its conceptual framework, test–retest reliability, and construct validity; inclusion in a longitudinal study is also required for analyses of the ability to detect change and to understand the individual-level responder definition of meaningful change. These upcoming steps will ensure that the SSM is fit-for-purpose as an efficacy endpoint in clinical trials evaluating SIBO treatments.