Biomarkers for diagnosis of acute appendicitis in adults

Objectives

This is a protocol for a Cochrane Review (diagnostic). The objectives are as follows:

To determine the diagnostic accuracy of biomarkers for detecting acute appendicitis in hospitalised adults with suspected acute appendicitis.

Secondary objectives

To determine (if possible) the diagnostic accuracy of biomarkers for detecting:

• uncomplicated and complicated (gangrenous or perforated) acute appendicitis;

• acute appendicitis in males and females;

• acute appendicitis in women of reproductive age;

• acute appendicitis in pregnancy;

• acute appendicitis at different time intervals from onset of pain.

Background

Target condition being diagnosed

Acute appendicitis remains a global disease with a lifetime incidence of 7% to 9% (Ferris 2017; Sellars 2017). Although the incidence of acute appendicitis has remained stable in western countries, it is increasing in newly industrialised countries (Ferris 2017). The pathogenesis of acute appendicitis is poorly understood but it is thought to be related to obstruction of the appendicular lumen by faecoliths, lymphoid tissue hyperplasia, tumours, foreign bodies, parasitic or fungal infections, inflammatory bowel disease, or due to trauma (Bhangu 2015). Variation in anatomical position of appendix, together with variation in clinical presentation of acute appendicitis, would make early and accurate diagnosis of appendicitis challenging (Bhangu 2015; Vieira 2019). Failure to diagnose acute appendicitis at an early stage can lead to morbidity and mortality due to complications such as perforation, abscess formation, and peritonitis. However, an incorrect diagnosis of acute appendicitis would lead to removal of a histologically normal appendix, which is known as a 'negative appendicectomy'. It is well recognised that negative appendectomies, comprising up to 20% of appendectomies, may result in increased risk of postoperative complications, length of hospital stay, and healthcare costs (Bijnen 2003; Lee 2014). It has been shown that the risk of a normal appendicectomy is higher in women (Bhangu 2020); this is thought to be due to the increased prevalence of alternative diagnoses that may mimic the presentation of appendicitis in women, including tubo‐ovarian pathologies, mid‐cycle pain, and pelvic inflammatory disease (Rothrock 1995).

Index test(s)

There are currently many blood biomarkers that are either being used in clinical practice, being validated in clinical studies, or being evaluated in experimental settings. These include C‐reactive protein (CRP), white cell count (WCC), neutrophil count (absolute value or percentage of WCC) (Andersson 2004), neutrophil‐to‐lymphocyte ratio (NLR) (Hajibandeh 2020), interleukin (IL)‐1 to IL‐10 (Paajanen 2002), bilirubin (Burcharth 2013), procalcitonin (Yu 2013), calprotectin (Thuijls 2011), pancreatic stone protein (Tschuor 2012), D‐lactate (Duzgun 2006), D‐dimer (Kaya 2012), fibrinogen (Nyuwi 2017), serum amyloid A (Schellekens 2013), mean platelet volume (Albayrak 2011), phospholipase A2 (Grönroos 1994), leukocyte elastase (Eriksson 1995), lactoferrin (Thuijls 2011), plasma total‐oxidant capacity (Ozdogan 2006), adenosine deaminase (Öztürk 2008), lipopolysaccharide binding protein (Brănescu 2012), and nuclear factor‐kappaB (Pennington 2000).

Clinical pathway

The main presenting complaint of people with acute appendicitis is abdominal pain, typically occurring in the right lower quadrant of the abdomen, which may be associated with nausea, vomiting, loss of appetite, fever, or a combination of these (Petroianu 2012). Variation in anatomical position of the appendix, together with variation in clinical presentation of acute appendicitis, would make early and accurate diagnosis of appendicitis challenging (Bhangu 2015; Vieira 2019). Therefore, diagnostic pathway involves history taking, clinical examination, and often a period of active observation to elicit progression of symptoms (Andersson 2004). This will be supplemented by biochemical markers of inflammation, clinical scoring systems, and imaging modalities.

Role of index test(s)

Biomarkers constitute an important part of the preoperative diagnostic work‐up in people with suspected acute appendicitis. Identifying a biomarker that can accurately diagnosis acute appendicitis and can differentiate between uncomplicated and complicated appendicitis with good sensitivity and specificity could inform the decision to perform imaging or operation without imaging and potentially reduce negative appendectomy rates. Moreover, the role of such a biomarker may be even more significant in assessing patients who do not routinely undergo computerised tomography (CT) scan (pregnant women or children) and in countries or settings where "twenty‐four seven" access to immediate imaging is limited (Hajibandeh 2020). In addition, a biomarker that could accurately distinguish between uncomplicated and complicated appendicitis has a potential to facilitate decision making in terms of prioritising the cases with clinically or radiologically confirmed appendicitis waiting for emergency appendectomy in busy emergency settings and in terms of monitoring people with radiologically confirmed uncomplicated appendicitis who are being treated conservatively (Hajibandeh 2020).

Alternative test(s)

Clinical scoring systems and imaging modalities are considered as alternatives to biomarkers in diagnosing acute appendicitis. The available clinical scoring systems are associated with suboptimal accuracy and specificity (Bhangu 2020), reducing their clinical usefulness in diagnosing acute appendicitis. Unlike clinical scoring systems, imaging modalities including ultrasonography (US), CT, and magnetic resonance imaging (MRI) play significant roles in the diagnostic pathway of acute appendicitis (Cobben 2009; Rud 2012; van Randen 2008). US has an estimated sensitivity of 78% and specificity of 83%, and its diagnostic accuracy is operator‐dependent and may be poor in adults compared to children due to increased adipose tissue with age (Doria 2006; van Randen 2008). CT with an estimated sensitivity of 91% and specificity of 90% has the advantage of being less operator‐dependent and its interpretation is not affected by bowel gas pattern, adipose tissue, or patient discomfort (van Randen 2008). Exposure to ionising radiation is considered one of the main disadvantages of CT, which is especially undesirable in young people with concerns of increasing lifetime risk of cancer (Brenner 2007). Moreover, CT is contraindicated in pregnancy. The use of MRI in the diagnosis of appendicitis is a relatively recent advance and is not widely implemented. MRI protocols for appendicitis and training for interpretation of scans are under development (Cobben 2009). However, MRI is currently recommended by the American College of Radiology in suspected appendicitis in pregnancy when an US examination is negative or inconclusive and where ionising radiation should be avoided (Leeuwenburgh 2012). Nevertheless, all of the aforementioned imaging modalities have cost implications and varying availability.

Rationale

Biomarkers constitute an important part of the preoperative diagnostic work‐up in people with suspected acute appendicitis; consequently, there are currently many biomarkers that are used in clinical practice, being validated in clinical studies, or being evaluated in experimental settings. Although the available imaging modalities have significantly improved the accuracy in diagnosis of acute appendicitis, the negative appendicectomy rate remains as high as 20% and increases to 30% in women of reproductive age (National Surgical Research Collaborative 2013). Identifying a biomarker that can accurately diagnosis acute appendicitis and can differentiate between uncomplicated and complicated appendicitis, together with clinical assessment and imaging modalities, could potentially decrease negative appendicectomy rates and could facilitate decision making:

• in people who do not routinely undergo CT scan (pregnant women or children);

• in settings where immediate access to imaging is limited;

• in prioritising the cases with clinically or radiologically confirmed appendicitis awaiting emergency appendectomy in busy emergency settings;

• in monitoring people with radiologically confirmed uncomplicated appendicitis who are being treated conservatively.

Objectives

To determine the diagnostic accuracy of biomarkers for detecting acute appendicitis in hospitalised adults with suspected acute appendicitis.

Secondary objectives

To determine (if possible) the diagnostic accuracy of biomarkers for detecting:

• uncomplicated and complicated (gangrenous or perforated) acute appendicitis;

• acute appendicitis in males and females;

• acute appendicitis in women of reproductive age;

• acute appendicitis in pregnancy;

• acute appendicitis at different time intervals from onset of pain.

Methods

Criteria for considering studies for this review

Types of studies

We will consider all single‐gate or two‐gate cross‐sectional clinical studies with either direct comparative (fully paired or unpaired) or indirect comparative design eligible for inclusion. We will consider studies that are incorrectly described as cohort (retrospective or prospective) by their original authors but are found to have cross‐sectional design eligible. We will exclude diagnostic case‐control studies, case reports, and reviews.

Participants

All hospitalised adults with clinically suspected acute appendicitis will be eligible for inclusion. Clinically suspected acute appendicitis comprises patients admitted to hospital with right lower quadrant abdominal pain in keeping with appendicitis. Adults are considered separately from children due to varying normal ranges of biomarkers between the two groups. The definition of adult varies in the literature from over 14 years to over 18 years of age. Therefore, we will exclude studies that include children younger than 14 years. It is important to evaluate the diagnostic test on the same population on which the test is applied in clinical practice as patient spectrum affects test performance characteristics (Bentley 2012). We will exclude studies conducted in immunocompromised people due to a theoretical alteration in biomarker response in this group. We will exclude populations from countries where neglected tropical diseases (NTDs) are endemic due to the differing prevalence of pathology in these regions, particularly an increased presentation of parasitic infections and tuberculosis (Appendix 1). Moreover, we will exclude studies specifically conducted on people with systemic conditions causing abdominal pain, such as porphyria, sickle cell anaemia, and familial Mediterranean fever.

Index tests

The following blood biomarkers will be index tests of interest: CRP, WCC, neutrophil count (absolute value or percentage of WCC), NLR, IL‐1 to IL‐10, bilirubin, procalcitonin, calprotectin, pancreatic stone protein, D‐lactate, D‐dimer, fibrinogen, serum amyloid A, mean platelet volume, phospholipase A2, leukocyte elastase, lactoferrin, plasma total‐oxidant capacity, adenosine deaminase, lipopolysaccharide binding protein, and nuclear factor‐kappaB. If combinations of biomarkers are reported in sufficient numbers, then each defined combination will be considered separately as an index test. Moreover, if serial measurements of biomarkers are combined in a defined manner in sufficient numbers, each defined combination will be considered as an index test.

Target conditions

Acute appendicitis will be the target condition.

Reference standards

The following reference standards will be used.

• Histological evidence of inflammation of the appendix will be the reference standard for operatively managed patients with excised appendix.

• Intraoperative evidence of inflammation of the appendix will be the reference standard for operatively managed patients with non‐excised appendix.

• Radiological evidence of inflammation of the appendix detected by CT or MRI will be the reference standard in non‐operatively managed patients.

Search methods for identification of studies

Electronic searches

We will develop comprehensive search strategy using appropriate keywords, thesaurus headings, search limits, and operators (Appendix 2; Appendix 3; Appendix 4; Appendix 5). Two review authors (SH, SH) with experience in evidence synthesis will independently apply and adopt the developed strategy on the following electronic databases: Cochrane Central Register of Controlled Trials (CENTRAL), MEDLINE Ovid, Embase Ovid, and CINAHL. There will be no language restrictions on the search strategy but it will be limited by date from 1990 to the current day in order to reduce heterogeneity in assays available for CRP.

Searching other resources

To identify potentially eligible unpublished or ongoing studies, we will explore the following sources: The European Association for Grey Literature Exploitation; System for Information on Grey Literature; World Health Organization International Clinical Trials Registry; International Standard Randomised Controlled Trial Number Registry; and ClinicalTrials.gov. Moreover, we will search relevant international surgical conference proceedings to identify abstracts. Finally, we will screen the reference lists of relevant reviews and articles to identify potentially eligible studies that are not identified by the above strategies.

Data collection and analysis

Selection of studies

To select eligible studies, two review authors (SH, SH) will independently screen the articles identified after application of the search strategies against the eligibility criteria. The process of screening will involve reading the titles and abstracts of the identified articles followed by full‐text retrieval and selection of potentially eligible studies. We will resolve disagreements between the first two review authors by discussion with a third review author (NJS).

Data extraction and management

We will use a random pilot‐testing technique to develop an online data extraction sheet consistent with Cochrane's data collection form. Two review authors (SH, SH) will independently extract the following data from the included studies: first author, country of origin, year and journal of the published study, study design, sample size, patient presentation, age, gender, pregnancy status, imaging modalities used, operative or non‐operative management, reference standards used, complicated or uncomplicated appendicitis, negative appendicectomies, biomarkers evaluated, and biomarkers diagnostic performance characteristics. We will resolve disagreements between the first two review authors by discussion with a third review author (NJS).

Assessment of methodological quality

Two review authors (SH, SH) will independently use the Quality Assessment of Diagnostic Accuracy Studies 2 (QUADAS‐2) criteria to assess the quality of the included studies (Whiting 2011). The precise criteria that will be used to make judgements about the risk of bias and problems with applicability are outlined in Appendix 6. We will resolve disagreements between the first two review authors by discussion with a third review author (NJS).

Statistical analysis and data synthesis

First, we will construct two‐by‐two tables for each index test based on the reported number of true positives, false negatives, true negatives, and false positives in included studies. Then, we will construct the coupled forest plots of the estimated sensitivities, specificities, and their 95% confidence intervals (CIs) and will plot them in the receiver operating characteristic (ROC) space for each index test using Review Manager 5 (Review Manager 2020). We will then assess the variability in estimated sensitivities and specificities for each index test across the included studies, taking into account the reported thresholds, characteristics and methodological quality of the included studies, and relevant covariates. We plan to estimate a summary sensitivity and specificity point at a chosen threshold for each index test using the bivariate logit normal random‐effects model (Reitsma 2005), when adequate data are provided for the chosen threshold by at least four studies. Moreover, we plan to estimate a summary receiver operating characteristic (SROC) curve for each test to take into account between‐study variations in thresholds using the hierarchical summary receiver operating characteristics (HSROC) model (Rutter 2001). We will use OpenMeta[Analyst] software for bivariate model and HSROC model analyses (OpenMeta[Analyst]). We are mindful of the new advanced approaches that are used to pool different thresholds in the same study and we will explore the use of methodology described by Jones and colleagues for better use of the available data in each study (Jones 2019). In this, the first round of this systematic review, we will be focusing on clarifying the accuracy of individual tests and combinations thereof where available. Given the scale of this task, we will not extend the review to consider comparisons of tests. A further justification for this is that there is currently no consensus about the most clinically relevant alternative diagnostic strategies which should sensibly be compared. We believe an initial review focused on individual tests and combinations will inform such a debate, and that comparing alternative biomarkers is premature.

Investigations of heterogeneity

The wider framework for investigation of heterogeneity will include aspects of the index test (specific manufacturer, version of test, and threshold (if no established threshold in clinical practice)), target disorder (proportion of simple or complicated appendicitis), target population (age and prior symptoms/investigation), and study quality (nature of reference standard). We will explore heterogeneity in the first instance through visual examination of forest plots of sensitivities and specificities, and through visual examination of the ROC plots. We will then include possible explanators of the heterogeneity as covariates in the HSROC model. Of the potential sources of heterogeneity listed, we think that the precise nature of the biomarker and the nature of the reference standard will be, a priori, the most important influences on heterogeneity, and so we will preferentially examine the influence of these where the number of included studies limits our ability to fully explore potential sources of heterogeneity.

Sensitivity analyses

Where appropriate, and if sufficient data are available, we will explore the sensitivity of any summary accuracy estimates to aspects of study quality using the number of domains where the QUADAS‐2 quality assessment identified risk of bias. Provisionally, we will consider the effect of excluding studies where the reference standard domain was at high or unclear risk of bias, as this is considered the most relevant source of bias. We will then consider the effect of excluding studies where any two or more domains (patient selection, index test, reference standard, flow and timing) were at high or unclear risk of bias.

Assessment of reporting bias

Quantitative methods for exploring reporting bias are not well established for studies of diagnostic test accuracy. Specifically, we will not consider funnel plots of the diagnostic odds ratio versus the standard error of this estimate.

What's new

Date	Event	Description
19 October 2021	New citation required and major changes	Shahab Hajibandeh and Shahin Hajibandeh joined the author team. The protocol has been rewritten in line with 2021 standards.

History

Protocol first published: Issue 3, 2015

Date	Event	Description
18 January 2021	Amended	Protocol updated

Appendices

Appendix 1. List of excluded populations from countries where neglected tropical diseases are endemic

Afghanistan, Algeria, American Samoa, Angola, Antigua and Barbuda, Argentina, Azerbaijan, The Bahamas, Bangladesh, Barbados, Belize, Benin, Bhutan, Bolivia, Botswana, Brazil, Brunei, Burkina Faso, Burundi, Cambodia, Cameroon, Cape Verde, Central African Republic, Chad, China, Colombia, Comoros, Democratic Republic of the Congo, Republic of the Congo, Cook Islands, Costa Rica, Cote d'Ivoire, Cuba, Djibouti, Dominica, Dominican Republic, Ecuador, Egypt, El Salvador, Equatorial Guinea, Eritrea, Ethiopia, Fiji, French Guiana, French Polynesia, Gabon, The Gambia, Ghana, Grenada, Guatemala, Guinea, Guinea‐Bissau, Guyana, Haiti, Honduras, India, Indonesia, Iran, Iraq, Jamaica, Kenya, Kiribati, North Korea, Kyrgyzstan, Laos, Lesotho, Liberia, Libya, Madagascar, Malawi, Malaysia, Maldives, Mali, Marshall Islands, Mauritania, Mauritius, Mexico, Federated States of Micronesia, Morocco, Mozambique, Myanmar (Burma), Namibia, Nauru, Nepal, New Caledonia, Nicaragua, Niger, Nigeria, Niue, Oman, Pakistan, Palau, Panama, Papua New Guinea, Peru, Philippines, Rwanda, Saint Kitts and Nevis, Saint Lucia, Saint Vincent and the Grenadines, Samoa, Sao Tome and Principe, Saudi Arabia, Senegal, Seychelles, Sierra Leone, Solomon Islands, Somalia, South Africa, Sri Lanka, Sudan, South Sudan, Suriname, Swaziland, Syria, Tajikistan, Tanzania, Thailand, Timor‐Leste, Togo, Tonga, Trinidad and Tobago, Turkey, Tuvalu, Uganda, Uruguay, Vanuatu, Venezuela, Vietnam, Wallis and Futuna, Yemen, Zambia, Zimbabwe (World Health Organization 2020).

Appendix 2. CENTRAL

#1 MeSH descriptor: [Appendicitis] explode all trees
#2 MeSH descriptor: [Appendectomy] explode all trees
#3 MeSH descriptor: [Appendix] explode all trees
#4 (appendectom* or appendice* or appendicit* or appendix):ti,ab,kw
#5 (#1 or #2 or #3 or #4)
#6 MeSH descriptor: [Biological Markers] explode all trees
#7 MeSH descriptor: [C‐Reactive Protein] explode all trees
#8 MeSH descriptor: [Leukocyte Count] explode all trees
#9 MeSH descriptor: [Interleukins] explode all trees
#10 MeSH descriptor: [Bilirubin] explode all trees
#11 MeSH descriptor: [Leukocyte L1 Antigen Complex] explode all trees
#12 MeSH descriptor: [Lithostathine] explode all trees
#13 MeSH descriptor: [Fibrinogen] explode all trees
#14 MeSH descriptor: [Mean Platelet Volume] explode all trees
#15 MeSH descriptor: [Phospholipases A2] explode all trees
#16 MeSH descriptor: [Phospholipases A2] explode all trees
#17 MeSH descriptor: [Leukocyte Elastase] explode all trees
#18 MeSH descriptor: [Lactoferrin] explode all trees
#19 MeSH descriptor: [Adenosine Deaminase] explode all trees
#20 MeSH descriptor: [NF‐kappa B] explode all trees 126
#21 (marker* or biomarker* or mutation* or C‐reactive protein or CRP or white cell count or WCC or leucocyt count or neutrophil count or IL‐1 or IL‐2 or IL‐3 or IL‐4 or IL‐5 or IL‐6 or IL‐7 or IL‐8 or IL‐9 or IL‐10 or bilirubin or procalcitonin or calprotectin or lithostathine or D‐lactate or D‐dimer or fibrinogen or serum amyloid or mean platelet volume or phospholipase A2 or leucocyte elastase or lactoferrin or plasma total‐oxidant capacity or adenosine deaminase or lipopolysaccharide binding protein or nuclear factor‐kappaB):ti,ab,kw
#22 (#6 or #7 or #8 or #9 or #10 or #11 or #12 or #13 or #14 or #15 or #16 or #17 or #18 or #19 or #20 or #21)
#23 (#5 and #22)

Appendix 3. MEDLINE (Ovid) search strategy

1. Appendicitis/
2. Appendectomy/
3. Appendix/
4. (appendec* or appendic* or appendix).tw.
5. 1 or 2 or 3 or 4
6. Biological Markers/
7. exp C‐Reactive Protein/
8. exp Leukocyte Count/
9. exp Interleukins/
10. exp Bilirubin/
11. exp Leukocyte L1 Antigen Complex/
12. exp Lithostathine/
13. exp Fibrinogen/
14. exp Mean Platelet Volume/
15. exp Phospholipases A2/
16. exp Leukocyte Elastase/
17. exp Lactoferrin/
18. exp Adenosine Deaminase/
19. exp NF‐kappa B/
20. (marker* or biomarker* or mutation* or C‐reactive protein or CRP or white cell count or WCC or leucocyt count or neutrophil count or IL‐1 or IL‐2 or IL‐3 or IL‐4 or IL‐5 or IL‐6 or IL‐7 or IL‐8 or IL‐9 or IL‐10 or bilirubin or procalcitonin or calprotectin or lithostathine or D‐lactate or D‐dimer or fibrinogen or serum amyloid or mean platelet volume or phospholipase A2 or leucocyte elastase or lactoferrin or plasma total‐oxidant capacity or adenosine deaminase or lipopolysaccharide binding protein or nuclear factor‐kappaB).tw.
21. 6 or 7 or 8 or 9 or 10 or 11 or 12 or 13 or 14 or 15 or 16 or 17 or 18 or 19 or 20
22. 5 and 21
23. limit 22 to yr="1990 ‐Current"

Appendix 4. Embase (Ovid) search strategy

1. exp appendicitis/
2. exp acute appendicitis/
3. exp appendectomy/
4. exp appendix/
5. (appendectom* or appendice* or appendicit* or appendix).tw.
6. 1 or 2 or 3 or 4 or 5
7. *biological marker/
8. *C reactive protein/
9. *leukocyte count/
10. *cytokine/
11. *bilirubin/
12. *calgranulin/
13. *lithostathine/
14. *fibrinogen/
15. *thrombocyte volume/
16. *phospholipase A2/
17. *leukocyte elastase/
18. *Lactoferrin/
19. *adenosine deaminase/
20. (marker* or biomarker* or mutation* or C‐reactive protein or CRP or white cell count or WCC or leucocyt count or neutrophil count or IL‐1 or IL‐2 or IL‐3 or IL‐4 or IL‐5 or IL‐6 or IL‐7 or IL‐8 or IL‐9 or IL‐10 or bilirubin or procalcitonin or calprotectin or lithostathine or D‐lactate or D‐dimer or fibrinogen or serum amyloid or mean platelet volume or phospholipase A2 or leucocyte elastase or lactoferrin or plasma total‐oxidant capacity or adenosine deaminase or lipopolysaccharide binding protein or nuclear factor‐kappaB).tw.
21. 7 or 8 or 9 or 10 or 11 or 12 or 13 or 14 or 15 or 16 or 17 or 18 or 19 or 20
22. 6 and 21
23. limit 22 to yr="1990 ‐Current"

Appendix 5. Science Citation Index‐Expanded/Conference Proceedings Citation Index‐Science search strategy

#1 TOPIC: ((appendectom* or appendice* or appendicit* or appendix))
#2 TOPIC: ((marker* or biomarker* or mutation* or C‐reactive protein or CRP or white cell count or WCC or leucocyt count or neutrophil count or IL‐1 or IL‐2 or IL‐3 or IL‐4 or IL‐5 or IL‐6 or IL‐7 or IL‐8 or IL‐9 or IL‐10 or bilirubin or procalcitonin or calprotectin or lithostathine or D‐lactate or D‐dimer or fibrinogen or serum amyloid or mean platelet volume or phospholipase A2 or leucocyte elastase or lactoferrin or plasma total‐oxidant capacity or adenosine deaminase or lipopolysaccharide binding protein or nuclear factor‐kappaB))
#3 (#1 AND #2)

Appendix 6. Review‐specific tailoring of QUADAS‐2

Domain 1: patient selection

Risk of bias: could the selection of participants have introduced bias?

Signalling questions and answering guidelines

1. Was a consecutive or a random sample of participants enrolled?

Answer 'yes’ if the study explicitly reports that a consecutive or random sample of participants was recruited. To fulfil this criterion, recruitment must have taken place at all hours and all days of the enrolment period.

Answer 'no' if the above statement was not fulfilled.

Answer 'unclear' if there is insufficient information available to answer 'yes' or 'no'.

2. Was a case‐control design avoided?

This question is irrelevant because case‐control studies are excluded from the review. The default answer would always be 'yes' raising no concerns of bias.

3. Did the study avoid inappropriate exclusions?

Answer 'yes' if the exclusion criteria would not be expected to alter the estimated diagnostic performance of the index test. Appropriate exclusions include people with anxiety, needle phobia, limited understanding, or communication, and people lacking capacity to consent.

Answer 'no' if the exclusion criteria would be expected to alter estimated diagnostic performance. For example, excluding 'difficult to diagnose' patients, such as women of reproductive age, or older people, would overestimate diagnostic accuracy.

Answer 'unclear' if there is insufficient information available to answer 'yes' or 'no'.

Guidelines for assessing risk of bias

Risk of bias from patient selection will be assessed as 'low' when the answer to signalling questions 1 and 3 is 'yes'.

Risk will be assessed as 'high' when the answer to either signalling question is 'no'.

Risk will be assessed as 'unclear' when there is insufficient information available to answer 'yes' to signalling questions 1 and 3.

Applicability: are there concerns that the included participants and setting do not match the review question?

Guidelines for assessing concern regarding applicability

Concerns regarding applicability relating to patient selection will be assessed as 'low' when the study population represents an unselected sample of adults with clinically suspected appendicitis. Appropriate exclusions are patients that do not match the review question, such as people who are immunocompromised where it would be reasonable to expect biomarker performance to be altered to the extent that estimated performance characteristics from a general population would not be applied to these patients in clinical practice. Pregnancy is likely to be excluded in some studies due to concerns regarding the influence of altered physiology on biomarker normal ranges. This would be considered as an acceptable exclusion, as although we intend to perform subgroup analysis on this group and have, therefore, not excluded them from the review, the results of biomarker performance in the general population are not likely to be applied to pregnant women in clinical practice. If inappropriate exclusions account for 5% or less of the included participants, the potential impact will be considered negligible.

Concern will be assessed as 'high' when the study population does not represent an unselected sample of adults with suspected appendicitis.

Concern will be assessed as 'unclear' when insufficient information is reported.

Domain 2: index test

Risk of bias: could the conduct or interpretation of the index test have introduced bias?

Signalling questions and answering guidelines

1. Were the index test results interpreted without knowledge of the results of the reference standard?

In most cases, the index test is an objective measurement carried out before the reference standard and is, therefore, not subject to bias. However, some experimental biomarker assays maybe carried out post hoc and may, therefore, be interpreted with knowledge of the reference standard.

Answer 'yes' if one of the following conditions is met:

a. the index test was carried out before surgery or patient discharge from hospital;

b. the biomarker assay is performed after the patient is discharged by a technician who is blinded to the reference standard, i.e. histology results or discharge diagnosis.

Answer 'no' if neither of the conditions are met.

Answer 'unclear' if there is insufficient information available to answer 'yes' or 'no'.

2. If a threshold was used, was it prespecified?

Answer 'yes' if a prespecified biomarker positivity threshold is stated.

Answer 'no' if it is stated that a threshold was not prespecified or that information in the article suggests it was not prespecified (e.g. non‐round numbers such as threshold of 12.3 ng/mL).

Answer 'unclear' if there is insufficient information available to answer 'yes' or 'no'.

Guidelines for assessing risk of bias

Risk of bias from index test will be assessed as 'low' when the answer to signalling questions 1 and 2 is 'yes'.

Risk will be assessed as 'high' when the answer to signalling questions 1 or 2 is 'no'.

Risk will be assessed as 'unclear' when insufficient information is available to answer signalling questions 1 and 2.

Applicability: are there concerns that the index test, its conduct, or its interpretation differ from the review question?

Guidelines for assessing concern regarding applicability

Biomarker assays must be standardised across eligible studies for each index test. Assays for the conventional biomarkers such as C‐reactive protein and white cell count are well described. Experimental biomarkers are likely to have greater variation in assay technique, and so the most widely performed assay technique for each index test will be identified from included papers and considered as the standard.

Concern regarding applicability relating to index test will be assessed as 'low' if one of the following conditions is met:

a. a standard assay is employed for conventional biomarkers;

b. the assays for experimental biomarkers are consistent with the most widely performed assay identified from included papers for each specific index test.

Concern will be assessed as 'high' if the biomarker assay does not fulfil either of the above conditions.

Concern will be assessed as 'unclear' if there is insufficient information available.

Domain 3: reference standard

Risk of bias: could the reference standard, its conduct, or its interpretation have introduced bias?

Signalling questions and answering guidelines

1. Is the reference standard likely to correctly classify the target condition?

Answer 'yes' if the diagnosis of appendicitis (or its absence) is based on one of the following criteria:

a. histological assessment of the excised appendix or surgical assessment of the appendix if it is deemed non‐inflamed and is, therefore, left in situ.

b. Computerised tomography (CT) diagnosis of non‐operatively managed patients who underwent CT imaging.

c. discharge diagnosis, readmission data, or clinical or telephone follow‐up in non‐operatively managed patients. The clinical reference standard must document resolution of symptoms at 7 to 31 days after the initial admission.

Answer 'no' if none of the above criteria are met.

Answer 'unclear' if there is insufficient information available to answer 'yes' or 'no'.

2. Were the reference standard results interpreted without knowledge of the results of the index test?

Answer 'yes' if all of the following conditions are met:

a. the pathologist performing histological assessment of the excised appendix is blinded to the index test result;

b. the surgeon performing the intraoperative assessment of an appendix that is to remain in situ is blinded to the index test result;

c. the radiologist interpreting the CT scan of a non‐operatively managed patient is blinded to the index test result;

d. the clinician or nurse assessing a patient on readmission or conducting clinical or telephone follow‐up is blinded to the index test result.

Answer 'no' if one of the above conditions is not met.

Answer 'unclear' if there is insufficient information available to answer 'yes' or 'no'.

Guidelines for assessing risk of bias

Risk of bias relating to the reference standard will be assessed as 'low' when the answer to signalling questions 1 and 2 is 'yes'.

Risk will be assessed as 'high' when the answer to signalling questions 1 or 2 is 'no'.

Risk will be assessed as 'unclear' when there is insufficient information available to answer signalling questions 1 and 2.

Applicability: are there concerns that the target condition as defined by the reference standard does not match the question?

Variations between the reference standard and the research question may arise from differing interpretations of histology or CT imaging. The accepted diagnosis of acute appendicitis on histology requires transmural inflammation with polymorphic granulocytes. The clinical importance of inflammation confined to the mucosa is dubious and would, therefore, not be applicable. The clinical reference standard is subject to the greatest degree of variations with the option for readmission data, clinic or telephone follow‐up which may all be carried out at differing time points over a period of 7 to 31 days following initial presentation. The details of the clinical reference standard will be extracted for descriptive purposes.

Domain 4: flow and timing

Risk of bias: could the patient flow have introduced bias?

Signalling questions and answering guidelines

1. Did all participants receive a reference standard?

Answer 'yes' if at least 95% of the included participants underwent a diagnosis by intraoperative, histological, or CT assessment or clinical follow‐up (30‐day readmission data, or clinical or telephone follow‐up).

Answer 'no' if less than 95% of the included participants underwent a diagnosis by intraoperative, histological, or CT assessment or clinical follow‐up (30‐day readmission data, or clinical or telephone follow‐up).

2. Did all participants receive the same reference standard?

We would not expect all participants to undergo surgery and it is, therefore, unlikely that all participants will have the same reference standard.

Answer 'yes' if one of the following conditions is met:

a. 90% of the included participants had surgery with histological examination of the excised appendix or intraoperative assessment of macroscopically normal non‐excised appendix;

b. 90% of the included participants with no surgery underwent clinical follow‐up (30‐day readmission, or clinic or telephone follow‐up within the 7‐ to 31‐day period following initial presentation).

Answer 'no' if neither of the above conditions is met.

Answer 'unclear' if there is insufficient information available to answer 'yes' or 'no'.

3. Were all participants included in the analysis?

Answer 'yes' if the analyses incorporated all participants. Also, answer 'yes' if 5% or less are excluded from the analysis because no reference standard assessment was available or if 5% or less were excluded due to missing or uninterpretable index test results, or both.

Answer 'no' if the above statement is not met.

Answer 'unclear' if insufficient information is available to answer 'yes' or 'no'.

Guidelines for assessing risk of bias

The risk of bias relating to patient flow and timing will be assessed as 'low' when the answer to signalling question 1, 2, and 3 is 'yes'.

The risk will be assessed as 'high' when the answer to signalling question 1, 2, or 3 is 'no'.

The risk will be assessed as 'unclear' when there is insufficient information available to answer signalling question 1, 2, or 3.

Contributions of authors

Task	Contributing author
Draft the protocol	SH, SH, MJM, NJS, IRD, CH, PW
Develop a search strategy	SH, SH, MJM, NJS, IRD
Search for studies (usually 2 people)	SH, SH
Obtain copies of studies Select which studies to include (2 + 1 arbiter) Extract data from studies (2 people) Enter data into Review Manager 5 Conduct the analysis Interpret the analysis Draft the final review Update the review	SH, SH SH, SH, NJS SH, SH SH, SH SH, SH, AMS, CH SH, SH, MJM, NJS, IRD, AMS, CH, PW SH, SH, MJM, NJS, IRD, AMS, PW SH, SH, MJM, NJS, IRD, AMS, CH

Sources of support

Internal sources

• Royal Devon and Exeter NHS Foundation Trust, UK

  Author salary, library and research facilities

• University of Exeter, UK

  Author salary, library and research facilities

• University of Exeter Medical School, UK

  Author salary and statistical support

External sources

• No sources of support provided

Declarations of interest

SH: none.

SH: none.

MJM: none.

NJS: none.

PGW: Involved in a study which may be eligible for inclusion via a Royal Devon & Exeter R&D small grant fund and the NIHR Exeter Clinical Research Facility.

CH: none.

AMS: Founder and Director of Attomarker Ltd, a biomarker detection platform. Attomarker Ltd did not fund the research.

IRD: none.

These authors contributed equally to this work.

These authors contributed equally to this work.

Amended to reflect a change in scope (see 'What's new')