Advice
Evidence review
Evidence review
Clinical and technical evidence
Regulatory bodies
A search of the Medicines and Healthcare products Regulatory Agency (MHRA) website revealed no manufacturer Field Safety Notices or Medical Device Alerts for this equipment.
No reports of adverse events were identified from searches of the Medicines and Healthcare products Regulatory Agency (MHRA) website, or from the US Food and Drug Administration database: Manufacturer and User Facility Device Experience (MAUDE).
Clinical evidence
Four fully published studies of the Peptest (Yuksel et al. 2012, Ocak et al. 2015, Hayat et al. 2015; Hayat et al. 2014) and 3 studies published as abstracts (Hayat et al. 2013, de Bortoli et al. 2012, de Bortoli et al. 2013) were identified to be relevant to this briefing. One of the fully published studies (Yuksel et al. 2012) was excluded because the manufacturer stated that it used an early prototype version of the Peptest and not the commercially available test. All of the fully published studies included adults only.
The study by Ocak et al. (2015, table 1) was a cross‑sectional study in Turkey, evaluating the accuracy of pepsin detection in saliva using the Peptest. The trial enrolled 20 patients with suspected laryngopharyngeal reflux (LPR) but the method of recruitment was not clear. All 20 had 24‑hour oesophageal pH monitoring. Each patient also gave a single sputum sample at the point in the 24 hours when they had their worst symptoms. In this study, the term 'pathologic gastro‑oesophageal reflux' was used to define the threshold at which gastro‑oesophageal reflux causes GORD.
Pathologic gastro‑oesophageal reflux was defined as the distal pH probe detecting pH of less than 4 for over 5% of the 24 hours. The Peptest results showed sensitivity of 33%, specificity of 100%, positive predictive value (PPV) of 100% and negative predictive value (NPV) of 14.2% for diagnosing pathologic gastro‑oesophageal reflux. The test accuracy of the Peptest for diagnosing LPR was not reported.
The UK‑based study by Hayat et al. (2015, table 2) had 2 aims: firstly, to establish normal values of salivary pepsin using the Peptest in healthy asymptomatic people, and secondly to determine the Peptest's ability to discriminate between people with reflux‑related symptoms, including GORD and hypersensitive oesophagus, and those with functional heartburn. A cut‑off value of 16 ng/ml pepsin in saliva was used to decide a positive test result for pepsin. Data were analysed from 111 patients with typical GORD symptoms (predominant heartburn with or without regurgitation) and 100 asymptomatic healthy volunteers. Both groups completed impedance‑pH (MII‑pH) monitoring and simultaneous salivary pepsin test on 3 samples during 24 hours. Outcome measures included prevalence of positive pepsin detection in saliva, pepsin concentration in saliva, timing of positive pepsin samples, correlation between pepsin in saliva and reflux parameters.
The study reported that, of the 111 symptomatic patients, 58 were objectively classified as having GORD, based on having an increased oesophageal acid exposure time greater than 4.2%, based on MII‑pH. The remaining 53 people had a normal oesophageal acid exposure time and were classified as either having hypersensitive oesophagus (n=26) or functional heartburn (n=27). Of the 58 objectively classified GORD patients, 45 had at least 1 positive test for pepsin out of the 3 samples taken. Twenty‑one of the patients classified as having hypersensitive oesophagus and 9 of those classified as having functional heartburn had at least 1 positive test for pepsin out of the 3 samples. Notably, 33 of the healthy, asymptomatic people had at least 1 positive test for pepsin out of 3 samples.
Hayat et al. (2014, table 3) aimed to quantify pharyngeal exposure to gastric contents using a number of new diagnostic techniques in patients with reflux‑related hoarseness and healthy people. The patient population in this study is therefore different from patients with GORD who do not have reflux‑related hoarseness. The study was conducted in the UK. It included 21 patients with hoarseness, who had been diagnosed with laryngopharyngeal reflux on the basis of clinical evaluation (questionnaires) and laryngoscopy, and 10 patients with no symptoms of GORD or voice disorders. Patients provided 5 saliva samples throughout the day to be analysed by the Peptest. The presence or absence of pepsin in the saliva was determined with a cut‑off value of 25 ng/ml.
The study found that there were 28 positive pepsin samples from 13 of the 21 patients with hoarseness (62%), and 6 positive samples from 4 of the 10 patients with no symptoms (40%). Only 1 of the 10 patients with no symptoms had 2 or more positive samples, compared with 9 of 21 with hoarseness.
Three studies that were published as abstracts were identified as being relevant to this briefing (Hayat et al. 2013, de Bortoli et al. 2012, de Bortoli et al. 2013). Two abstracts (Strugala et al. 2007a, 2007b) were also identified but were excluded, because the manufacturer stated that the Peptest used in these 2 studies was an early pre‑production, non‑CE‑marked prototype. The 3 relevant abstracts provided very limited information on aspects of the study settings, methods, characteristics and results. The studies are outlined in table 4.
Table 1 Summary of the Ocak et al. (2015) study
|
Study component |
Description |
|
Objectives/hypotheses |
To evaluate the accuracy of pepsin detection in the saliva using the Peptest for the diagnosis of LPR. |
|
Study design |
Cross‑sectional study. All patients had 24‑hour oesophageal dual pH monitoring, during which each patient gave 1 sample of sputum Peptest when they had the worst symptoms. The Peptest was expected to have the ability to detect pepsin down to 16 ng/ml. Ambulatory pH monitoring and pepsin detection test analysis were double‑blinded by separate researchers. |
|
Setting |
Unclear about the setting but the study was conducted in Turkey. Further details were not reported. No information was provided on the source of the patients, including how they were recruited. |
|
Inclusion/exclusion criteria |
Inclusion criteria: people with suspected LPR, who had at least 1 LPR symptom, with RSI>15 and RFS>3. Exclusion criteria: psychiatric disorders with cooperation disability; previous laryngeal surgery history; any kind of nasal, paranasal, pharyngeal, laryngeal or pulmonary disease which can mimic LPR symptoms; patients who had taken proton pump inhibitors in the last 1 month. |
|
Primary outcomes |
Pathologic GOR findings (percentage of time pH<4 in distal probe over 5%); LPR findings (presence of a single attack of pH<4 in the proximal probe); pH score in the proximal and distal probes when the sputum sample was given; sensitivity, specificity; PPV; NPV. |
|
Statistical methods |
Not reported. |
|
Participants |
People with a suspicion of LPR (n=20). |
|
Results |
Mean RSI: 22.1 Mean RFS: 8.1 Patients with at least 1 LPR attack (pH<4 in the proximal probe): 90%. Mean pH value: 6.38 in the proximal probe and 4.32 in the distal probe at the sample time. Patients with pathologic GOR findings (percentage of time pH<4 in distal probe over 5%): 90%. All pepsin‑positive patients were in the pathologic GOR group. For pepsin detection test for GOR:
Among the 6 people with a pepsin positive test, pH<4 in 66% of the distal probe and 33% of the proximal probe at the sampling time. Pepsin‑positive people had an apparent acidic pH value in the proximal probe at the sample time compared with the pepsin‑negative patients (3.26 compared with 6.81). |
|
Conclusions |
The authors concluded that, because of the benefits and ease of application, a positive salivary pepsin test in a patient suspected of having LPR can be a cost effective, accurate and alternative diagnostic method. Increasing the daily number of sputum samples may increase the sensitivity of the test. |
|
Abbreviations: GOR, gastro‑oesophageal reflux; LPR, laryngo‑pharyngeal reflux; n, number of patients; NPV, negative predictive value; PPV, positive predictive value; RFS, reflux finding score (details available in table 2 of the paper); RSI, reflux symptom index (0=no problem; 5=severe problem). |
|
Table 2 Summary of the Hayat et al. (2015) study
|
Study component |
Description |
|
Objectives/hypotheses |
To establish normal values of salivary pepsin in healthy asymptomatic people and to determine its value to discriminate patients with reflux‑related symptoms (GORD, HO) from functional heartburn, using Peptest. |
|
Study design |
A prospective controlled cohort study. Patient sampling was consecutive. Cut‑off pepsin concentration for a saliva sample positive for pepsin was >16 ng/ml. Each patient had 3 samples during the day: on waking, 1 hour after finishing lunch and 1 hour after finishing dinner during the 24‑hour ambulatory MII‑pH monitoring period. Analysis was blinded to the patients' status, any reflux monitoring parameter and Reflux Disease Questionnaire scores. |
|
Setting |
Healthy volunteers were recruited by advertisements placed at St George's University of London, and consecutive patients with typical GORD symptoms were referred to the Upper Gastrointestinal Physiology Unit at the Royal London Hospital for reflux assessment. |
|
Inclusion/exclusion criteria |
Patients with typical GORD symptoms (predominant heartburn with or without regurgitation) or with a primary complaint of heartburn were included. Patients were excluded if they had a history of previous oesophageal/gastric surgery, or a known oesophageal motor disorder (e.g. achalasia, scleroderma). For the control: asymptomatic healthy volunteers. |
|
Primary outcomes |
Prevalence of positive pepsin detection in saliva; pepsin concentration in saliva; timing of positive pepsin samples; correlation between pepsin in saliva and reflux parameters; pepsin concentration in saliva to differentiate patients with GORD, or patients with reflux‑related symptom (GORD+HO) from patients with FH. |
|
Statistical methods |
Data were expressed as mean±SEM or median (IQR) where appropriate. A p value of <0.05 was considered significant. A ROC curve was constructed to determine and compare the sensitivity and specificity of different pepsin cut‑off concentrations and their predictive value to diagnose or refute the diagnosis of GORD and reflux‑related symptoms. |
|
Participants |
|
|
Resultsc |
Of the 58 patients classified as having GORD, 45 had at least 1 positive test for pepsin out of the 3 samples. Twenty‑one of the patients classified as HO and 9 of those classified as FH had at least 1 positive test for pepsin out of the 3 samples. Diagnostic accuracy of Peptest in these 111 symptomatic patients was not reported in the paper but can be calculated as follows (based on number of patients tested): sensitivity 78%, specificity 43%, PPV 60%, NNP 64%, LR for positive test of 1.37, and LR for negative test 1.94, when at least 1 out of 3 samples was positive. Of the 87 asymptomatic healthy people who had a normal MII‑pH, 33 had at least 1 positive test for pepsin out of the 3 samples. Diagnostic accuracy of Peptest in the 58 patients objectively diagnosed with GORD comparing with these 100 healthy controls was not reported; it can be calculated as follows (based on number of patients tested): sensitivity 78%, specificity 62%, LR for positive test 2.05, LR for negative test 2.77, and false positive rate 38%, when at least 1 out of 3 samples was positive. However, this calculation presumes that all the controls had been confirmed to be truly negative by the same reference standard used for the patients. When measuring the highest pepsin concentration in saliva out of the 3 samples, 1 third of asymptomatic patients had pepsin in saliva at low concentration: median 0 ng/mL; 25–75th centiles 0–59; 95th centile 190.6. |
|
Conclusions |
In patients with symptoms suggestive of GORD, salivary pepsin testing may complement questionnaires to assist office‑based diagnosis. This may lessen the use of unnecessary anti‑reflux therapy and the need for further invasive and expensive diagnostic methods. |
|
Abbreviations: AET, (oesophageal) acid exposure time; ANOVA, analysis of variance; CI, confidence interval; FH, functional heartburn; GORD, gastro‑oesophageal reflux disease; HO, hypersensitive oesophagus; ITT, intention to treat; IQR, inter‑quartile range; LR, likelihood ratio; MII‑pH, impedance‑pH; NPV, negative predictive value; n, number of patients; PPV, positive predictive value; RR, relative risk; ROC, receiver operator characteristic curves; SEM, standard error of the mean. aSAP, symptom association probability, which was used to characterise the association between reflux and symptoms. No further details were described. bIt was unclear whether MII‑pH monitoring result for this group was also evaluated based on the cut‑off value of the oesophageal acid exposure time of 4.2%. cOnly outcomes that are relevant to this briefing report were reported in this table. |
|
Table 3 Summary of the Hayat et al. (2014) study
|
Study component |
Description |
|
Objectives/hypotheses |
To assess new methods for the objective detection of oesophago‑pharyngeal reflux and for quantification of pharyngeal exposure to gastric contents in patients with hoarseness and asymptomatic controls. |
|
Study design |
Case‑control study. Pepsin in the saliva sample was detected using Peptest. Each person had 5 saliva samples collected for the analysis of pepsin presence. The presence or absence of pepsin in the saliva was determined with a cut‑off value of 25 ng/ml. It was unclear whether the individual who tested the samples and who interpreted the test were blinded to sample source and subject classification (GORD compared with healthy volunteers). |
|
Setting |
Patients were recruited from a tertiary referral specialist voice clinic in the UK. No further details about the setting were reported, nor information about the source of healthy volunteers. |
|
Inclusion/exclusion criteria |
Patients with hoarseness who had been diagnosed with laryngo‑pharyngeal reflux on the basis of clinical evaluation (questionnaires) and laryngoscopy were included. Controls were healthy adult volunteers with no symptoms of GORD or voice disorders (evaluated in the same way with identical questionnaires and laryngoscopy). People who smoke were excluded, along with patients with chronic cough, significant pulmonary or neuro‑musculo‑skeletal disease, or where voice misuse was suspected. Patients were excluded if they had a previous laryngeal surgery, history of oro‑pharyngeal/laryngeal cancer, previous gastric or oesophageal surgery, or endotracheal intubation in the last 3 months. All patients were studied 'off' proton pump inhibitor treatment (at least 7 days). |
|
Primary outcomes |
Including saliva pepsin concentration, pH monitoring, GORD symptom questionnaire, Reflux Symptom Index, Voice Handicap Index, and Reflux Finding Score. |
|
Statistical methods |
Descriptive statistics are presented as mean±SEM for parametric data and median values with range where appropriate. The independent samples t test or Mann‑Whitney U test was used to compare median values. The Wilcoxon signed‑rank test was used to compare repeat questionnaire scores for patients. Correlation was performed using the Pearson test. The Fisher exact test was used for proportional differences. A p‑value of <0.05 was considered significant. |
|
Participants |
Patients (21) with hoarseness and 'positive' laryngoscopy who had been diagnosed with laryngo‑pharyngeal reflux on the basis of clinical evaluation (questionnaires) and laryngoscopy were included. Controls (n=10), adults with no symptoms of GORD or voice disorders. |
|
Resultsa |
5 samples from each subject were analysed. In the patient group, there were 28 positive pepsin samples from 13 of the 21 patients (62%). In the control group, there were 6 positive samples from 4 of the 10 controls (40%). Only 1 of the 10 controls had ≥2 samples positive for pepsin compared with 9 of the 21 patients. The saliva samples taken from patients were more likely to be positive (p=0.025). Diagnostic accuracy of Peptest comparing the patients with the controls was not reported but it can be derived as the following based on the above data:
The above calculations presume that all the controls had been confirmed to be true negative by the same reference standard used for the patients. |
|
Conclusions |
The authors concluded that 'a subgroup of patients with hoarseness (10/21) had objective detection of the oesophago‑pharyngeal reflux. We propose that these patients are more likely to benefit from further intense anti‑reflux therapy. Detection of pepsin in the saliva may be a useful screening tool in these patients.' |
|
Abbreviations: GORD, gastro‑oesophageal reflux disease; LR, likelihood ratio; n, number of patients; SEM, standard error of the mean. aOnly those outcomes that are relevant to this briefing were reported in this table. |
|
Table 4 Summary of abstracts
|
Study component |
Description |
|
Hayat et al. (2013) |
|
|
Objectives/hypotheses |
To measure pepsin in saliva with objective assessment of GORD by MII‑pH in a cohort of asymptomatic patients and consecutive patients with clinically significant heartburn (according to the Montreal definition of GORD). |
|
Tests |
Index test: Peptest for pepsin in saliva (the cut‑off value to determine pepsin positivity was 25 ng/ml). Reference test: MII‑pH monitoring. |
|
Participants |
65 consecutive patients with clinically significant heartburn and 100 healthy people. Patients were divided into 3 phenotypes based on MII‑pH results:
|
|
Results |
All healthy people selected had normal MII‑pH testing. 36/100 healthy patients had at least 1 sample positive (20% had 1 sample positive, 12% had 2 samples positive and 4% had 3 samples positive). In pepsin positive samples, the median (25%>75%) pepsin concentration was 118 (64–181) ng/ml. In GORD patients, 21/26 had at least 1 sample positive (3 patients had 3 samples positive) and pepsin concentration was 152 (72–250) ng/ml. In HO patients 15/18 had at least 1 sample positive (4 had 3 samples positive) and pepsin conc. was 250 (74–250) ng/ml. In contrast, only 2/12 FH patients had at least 1 sample positive and pepsin concentration was 76 (67–85) ng/ml. Peptest had a sensitivity of 95%, specificity 89%, PPV 97% and NPV of 57% for identifying patients with heartburn related to reflux (GORD+HO). |
|
de Bortoli et al. (2013) |
|
|
Objectives/hypotheses |
To evaluate the Peptest accuracy for the diagnosis of GORD in patients with reflux symptoms by means of MII‑pH. |
|
Tests |
Index test: Peptest for pepsin in saliva/sputum (cut‑off value for positive not stated) Reference test: MII‑pH. |
|
Participants |
35 patients with GORD symptoms were studied. All patients with negative endoscopy underwent patho‑physiological examinations. Patients were grouped on the basis of MII‑pH results as follows:
|
|
Results |
Peptest was positive in 93.7% of true‑GORD, in 58.3% of HE, and negative in 100% of no‑GORD patients. The accuracy of Peptest: sensitivity 100%; specificity 79%; positive predictive value 100%; negative predictive value 54%; diagnostic accuracy 83%. |
|
de Bortoli et al. (2012) |
|
|
Objectives/hypotheses |
To evaluate the Peptest accuracy by means of MII‑pH for the diagnosis of GORD in patients with reflux symptoms. |
|
Tests |
Index test: Peptest for pepsin in saliva/sputum (cut‑off value for positive not stated). Reference test: oesophageal manometry and MII‑pH. |
|
Participants |
26 patients with GORD symptoms (all underwent upper endoscopy and, if negative, patho‑physiological oesophageal examinations). Patients were grouped on the basis of MII‑pH results as follows:
|
|
Results |
Peptest was positive in 82% of true‑GORD, in 57% of HO, and negative in 100% of no‑GORD patients. It showed 72% sensitivity, 100% specificity, 100% positive predictive value, 62% negative predictive value, and 81% diagnostic accuracy. |
|
Abbreviations: AET, acid exposure time; GORD, gastro‑oesophageal reflux disease; HO, hypersensitive oesophagus; LR, likelihood ratio; MII‑pH, impedance‑pH; NPV, negative predictor value; n, number of patients; PPV, positive predictor value; RSI, reflux symptom index; SAP, symptom association probability index; SEM, standard error of the mean. |
|
Recent and ongoing studies
One ongoing or in‑development trial on the Peptest was identified in the preparation of this briefing:
-
NCT02183961: Comparison of 3 methods used in the diagnosis of extraesophageal reflux in children with chronic otitis media with effusion (enrolment: 24; study start date: June 2012; study completion date: May 2014).
Costs and resource consequences
If patients provide 3 Peptest samples at different times of day, as recommended by the manufacturer, the cost per test is £60 if the processing, testing and assessment is carried out by RD Biomed, or between £30 and £54 if processed and tested locally.
According to the National Schedule of Reference Costs for England, over 2011–12 there were 164,679 diagnostic endoscopic procedures on the upper gastro‑intestinal tract in patients aged 19 years and older, at a total cost of £71,829,933 with an average unit cost of £436.
The Peptest may provide a cost saving if a positive test result replaced the need for an endoscopic procedure. However, this will not be the case for all endoscopy referrals. The total potential cost savings would depend on the proportion of current endoscopy referrals that could be avoided where endoscopy was not indicated for other reasons, the proportion of those patients with acid reflux and the sensitivity of the test. A negative test result, either because the patient does not have GORD or the Peptest failed to diagnose it, would put the patient back on the pathway to current diagnoses using endoscopy or pH monitoring. In this case, the Peptest would be an additional care cost for these people.
Using Peptest could replace ambulatory pH monitoring where this test is carried out. According to NHS Tariff costs 2014–15 Disorders of the oesophagus with a length of stay of 1 day or less (FZ31F) has a unit cost of £392. It is unclear from the evidence how many such cases may be avoided by the use of Peptest.
Strengths and limitations of the evidence
In order to assess the usefulness of a new diagnostic test in clinical practice, evidence is needed to show:
-
how well the test works compared to a reference diagnostic test that is known to work well in clinical practice
-
the diagnostic yield of the test when it is introduced into a clinical pathway with other diagnostic tests
-
the therapeutic yield associated with the test, that is, the benefits in terms of patient outcomes from the introduction of the test.
Current evidence on the Peptest only partially addresses the first of these questions.
In the Ocak et al. (2015) study, pH monitoring analysis and pepsin detection test analysis were double‑blinded by separate researchers. However, the study was a very small cross‑sectional study of 20 people with laryngopharyngeal reflux, and the recruitment method was unclear. Therefore, it is unclear whether the people in the study were representative of the population for whom the Peptest is intended, or whether the setting was relevant to current NHS practice.
As the authors stated, pH monitoring is an imperfect reference standard. Furthermore only 1 sample was taken from each patient and this does not reflect how Peptest would be used in practice.
Hayat et al. (2015) was conducted in the UK and so the results may be more relevant to the NHS. The study had a relatively large sample size of 238 patients recruited consecutively, of whom 111 had symptoms of GORD and 100 did not. Three saliva samples were taken from each patient. The test accuracy was not included in the study report but has been calculated based on a subset of the patients in the study and their samples. Analysis was performed blinded to the patients' status, any reflux monitoring parameter and Reflux Disease Questionnaire scores. The test population consisted of patients with some primary symptoms of GORD. The publication did not state whether the included patients were people whose GORD symptoms had already failed to respond to proton pump inhibitors prior to their inclusion in the study. It was therefore unclear whether these patients matched the patients in whom Peptest might be used in clinical practice, because in practice it may be offered to people who have typical GORD symptoms and whose GORD had failed to respond to proton pump inhibitors and lifestyle changes.
The Hayat et al. (2014) study was also conducted in the UK, meaning the results may be more relevant to the NHS. It was a further small study including 31 patients. As with the Ocak et al. (2015) study, it was unclear how the patient sampling was done. It was also unclear whether the technicians were blinded to the status of samples, and to the order and timescale of testing.
The inclusion and exclusion criteria were somewhat different between the 3 studies, resulting in different populations and implications for interpretation of results. The various reference standards used in the studies, including MII‑pH results and endoscopy examinations, are imperfect in classifying GORD.
A range of salivary pepsin concentrations were used as cut‑off values for pepsin positivity in the samples across the studies, from greater than 16 ng/ml to more than 25 ng/ml. This makes it difficult to compare the results across the studies. There is no consensus in the published literature about standard and clinically relevant concentrations of pepsin in saliva. There was no comparable test for salivary pepsin, although laboratory tests were available and used primarily in a research context.
Between 1 and 5 tests were done per patient in different studies, and the calculated sensitivity and specificity are based on the number of patients tested. Pepsin positivity varied by time of day and proximity to eating. Furthermore, the 3 studies published as abstracts provided only very limited information in terms of study setting, methods, characteristics and results.
Overall, published evidence on the diagnostic accuracy of the Peptest for GORD and laryngopharyngeal reflux is of limited quantity and relevance. No published evidence on test characteristics for diagnosing GORD or laryngopharyngeal reflux from saliva or sputum in children was identified. No published evidence or useful information was identified to inform a clear role of the Peptest in a specific clinical pathway.