Faecal occult blood testing: a review of its use and common misutilisation

Abstract

Colorectal cancer (CRC) remains a leading cause of cancer-related morbidity and mortality worldwide, with early detection being crucial for improving patient outcomes. While colonoscopy is the gold standard for CRC screening, stool-based tests such as guaiac-based faecal occult blood test and faecal immunochemical test offer non-invasive and cost-effective alternatives. These tests have proven value in the outpatient screening of asymptomatic, average-risk individuals; however, their frequent misuse in inpatient settings limits their diagnostic accuracy and utility. Inappropriate use of faecal occult blood tests (FOBTs) in hospitalised patients often results in false-positive or false-negative findings, leading to unnecessary diagnostic procedures, delayed treatment, increased healthcare costs and potential patient harm. Addressing this issue requires promoting adherence to guideline-based use of FOBT, alongside targeted provider education to reduce misuse, improve diagnostic precision and optimise patient care.

Summary box.

• Stool-based FOBTs (gFOBT/FIT) are often misused in hospitalised patients, reducing diagnostic accuracy compared to their intended use in outpatient colorectal cancer screening.

• Inappropriate inpatient testing can lead to false results, unnecessary procedures, delayed care, increased costs, and potential patient harm.

• Adherence to established guidelines is essential to preserve test utility and improve clinical decision-making.

• Targeted provider education and system-level interventions can help reduce misuse and optimise patient outcomes.

Introduction

Colorectal cancer (CRC) is the third most common cancer worldwide, and early detection significantly improves survival outcomes.¹ CRC screening programmes have been implemented globally, with colonoscopy recognised as the gold standard for screening.² However, stool-based tests provide a non-invasive alternative and are frequently used in a two-step approach. This approach begins with a guaiac-based faecal occult blood test (gFOBT), faecal immunochemical test (FIT) or stool DNA test (eg, Cologuard), with positive results prompting follow-up colonoscopy to confirm the diagnosis or detect other abnormalities. These stool-based tests are both cost-effective and easy to administer, which has supported their widespread use, also raises concerns about potential misuse or inappropriate application in certain settings.

Numerous studies from both academic and community healthcare settings have highlighted the misutilisation of faecal occult blood tests (FOBTs), particularly as adjunctive tools in the evaluation of gastrointestinal (GI) bleeding.3,10 Despite limited diagnostic value in acute clinical presentations, FOBTs are frequently ordered during the workup for anaemia, overt GI bleeding, and abdominal pain. Even in CRC screening, strict protocols, including dietary and medication restrictions, must be followed to ensure test accuracy.

In the inpatient setting, evidence consistently demonstrates the limited utility of FOBTs, especially in cases of overt GI bleeding. However, misuse remains prevalent, leading to unnecessary testing, overuse of diagnostic procedures such as endoscopies and colonoscopies, delays in appropriate care and, ultimately, suboptimal patient outcomes. This review aims to: (1) summarise the appropriate role of FOBT in CRC screening based on current guidelines, (2) examine patterns of misuse and associated clinical consequences and (3) review strategies to reduce inappropriate utilisation.

Background and types of FOBT

FOBT was developed as a non-invasive, low-cost alternative for CRC screening as this is one of the most prevalent cancers in both men and women worldwide. The primary goal of FOBT is to reduce CRC-related mortality.¹¹ Screening for CRC is recommended by the United States Preventive Service Task Force starting at the age of 45 years in the average-risk population and can be done through colonoscopy, stool-based tests or CT colonography.¹² FOBT is particularly popular due to its simplicity, affordability and ease of administration, requiring no specialised equipment and offering results that can guide further diagnostic evaluation. The two most common types of FOBT are gFOBT and FIT.

The gFOBT detects occult blood through a peroxidase reaction. In this reaction, hydrogen peroxide catalyses the oxidation of guaiac, producing a blue colour if haemoglobin is present. The test is performed using a card with guaiac-impregnated paper. When a stool sample is applied to the paper and a hydrogen peroxide developer solution is added, the haematin portion of the haemoglobin (Hb) catalyses the release of oxygen. This released oxygen oxidises the guaiac, resulting in a blue colour change that indicates the presence of blood.

Before performing a gFOBT, patients should receive detailed instructions on proper preparation. They must avoid consuming red meat, certain vegetables and specific medications for 2 to 7 days before the test. False-positive results can occur due to the presence of animal-derived haemoglobin, consumption of green vegetables with chlorophyll-mediated pseudo-peroxidase activity or naturally high levels of peroxidase in some foods. These foods include but are not limited to broccoli, cauliflower, cantaloupe, carrots, grapefruit, melons and tomatoes. Medications that should be avoided to prevent false positives are acetylsalicylic acid, unfractionated or low-molecular-weight heparin, warfarin, clopidogrel, non-steroidal anti-inflammatory drugs (NSAIDs), selective serotonin reuptake inhibitor and iron supplements.¹³ Additionally, ascorbic acid (vitamin C) can also lead to false-negative results because it acts as a reducing agent. Proper sample collection is also critical. The test requires three consecutive stool samples. Collecting samples during a digital rectal exam (DRE), menstruation or when haemorrhoids are present is discouraged, as these situations increase the risk of false-positive results.¹⁴

Studies have shown that screening with gFOBT reduces CRC mortality. In a meta-analysis conducted by Zhang et al, a subanalysis of 19 studies including 2 264 603 participants comparing gFOBT to no screening showed a 14% reduction in CRC mortality (RR, 0.86; 95% CI 0.82 to 0.90).¹⁵ As summarised in table 1, gFOBT demonstrates moderate sensitivity (50%–70%) and high specificity (85%–95%) for CRC detection, though its sensitivity is significantly lower for detecting advanced adenomas and polyps. In addition, gFOBTs have become increasingly replaced by FIT in population-based CRC screening programmes due to their superior test performance.¹¹

Table 1. Sensitivity and specificity of stool-based tests for colorectal cancer (CRC), advanced adenomas and polyps.

Test	Target lesion	Sensitivity (%)	Specificity (%)
FOBT (Guaiac-based)	CRC	~50–70%	~85–95%
	Advanced adenomas	~10–30%	~85–95%
	Polyps	Low (~10%)	High (~95%)
FIT	CRC	~70–85%	~90–95%
	Advanced adenomas	~25–40%	~90–95%
	Polyps	Low (~20%)	High (~95%)

FIT, faecal immunochemical test; FOBT, faecal occult blood test.

The FIT is an immunoassay specifically designed to detect human haemoglobin in stool samples using antibodies that target either haemoglobin alone or in combination with the haemoglobin/haptoglobin (Hb/Hp) complex. This allows for the detection of even very low levels of faecal blood. The test is conducted by collecting a stool sample in a specially designed container, without requiring any prior dietary or medication restrictions. Unlike gFOBT, FIT is not affected by foods with peroxidase activity, eliminating the risk of false positives from dietary sources. Additionally, discontinuation of NSAIDs is generally unnecessary although the use of proton-pump inhibitors has been found to increase FIT positivity at the expense of false-positive results.¹⁴

Several studies have demonstrated FIT’s superior diagnostic performance over gFOBT. A randomised controlled trial by van Rossum et al demonstrated the superiority of FIT over gFOBT for detecting colorectal neoplasia, and a meta-analysis by Lee et al reported pooled sensitivity and specificity for CRC of 79% and 94%, respectively.^{16 17} In a head-to-head clinical trial, ColonView FIT outperformed gFOBT in sensitivity (95% vs 76%) while maintaining acceptable specificity (65% vs 84%). These findings support European guidelines that now recommend FIT over gFOBT for population-based CRC screening.¹¹

To further assess the diagnostic accuracy of these tests, Meklin et al conducted a comprehensive systematic review and meta-analysis of 31 studies from China, Europe and the USA. Focusing on single-round testing and using invasive CRC as the reference standard, the study found that FIT had superior sensitivity (86%) and comparable specificity (85%) compared with gFOBT (68% sensitivity, 88% specificity).¹⁸ Hierarchical summary receiver operating characteristic analysis confirmed the diagnostic advantage of FIT, with a significantly higher area under the curve (AUC: 0.87 vs 0.77; p=0.0017). Although many FIT brands are commercially available, no individual brand showed significantly better diagnostic performance. Both qualitative and quantitative FITs performed well, with Magstream, ColonView, InstantView and Prevent ID among the top-rated options.

Despite FIT’s strong analytical profile, global evidence regarding its impact on CRC mortality remains limited. The International Agency for Research on Cancer has endorsed biennial FIT screening based on modelled and extrapolated data, largely from gFOBT trials, as randomised controlled trials specifically assessing FIT’s long-term mortality benefit are lacking. Real-world data from Finland illustrate these challenges. A gFOBT-based national pilot programme (2004–2012) failed to demonstrate mortality reduction, leading to the adoption of an FIT-based screening programme in 2022. This programme currently relies on a single-sample FIT targeting Hb alone. However, a validation study comparing various strategies found that while one-sample Hb testing was effective for detecting invasive CRC (sensitivity ~94.5%), it performed poorly for adenomas (sensitivity ~12.5%).¹⁹ In contrast, one-sample testing using both Hb and Hb/Hp markers outperformed three-sample Hb-only testing. The highest diagnostic accuracy was achieved with two-sample FIT testing for both Hb and Hb/Hp complex, yielding a sensitivity of 47.5% and an AUC of 0.73—more than two times the sensitivity of the standard three-sample Hb-only approach. These findings highlight the critical importance of optimising both sampling strategy and biomarker selection in FIT-based screening programmes to improve the detection of CRC precursor lesions and maximise preventive effectiveness.

Current guidelines

According to current guidelines from various professional societies (table 2), FOBT is an approved screening tool for individuals aged 45 and older who are at average risk for CRC.1220,23 These guidelines emphasise that patients with positive FOBT results should undergo follow-up colonoscopy for further evaluation and accurate diagnosis.

Table 2. Overview of different societies’ guidelines for FOBT.

Society	Patient population	Frequency	Indication
USPSTF	Adults 45 years and older who do not have signs or symptoms of CRC and who are at average risk of CRC.	Every year	CRC screening
ACS	Adults 45 years and older with an average risk of CRC.	Every year	CRC screening
ACG	Recommend average-risk individuals between ages 50 and 75 years. Suggest average-risk individuals between ages 45 and 49.	Every year	CRC screening
NCCN	Adults 45 years and older with an average risk of CRC.	Every year	CRC screening
ASGE*	People 45 or older without prior colorectal cancer or polyps and without any of the factors that define high-risk screening.	Every year	CRC screening

^*Only mention faecal immunochemical test.

ACG, American College of Gastroenterology; ACS, American Cancer Society; ASGE, American Society of Gastrointestinal Endoscopy; CRC, colorectal cancer; NCCN, National Comprehensive Cancer Network; USPSTF, U.S. Preventive Services Task Force.

Notably, there have been no recent publications recommending or evaluating the use of FOBT outside the context of CRC screening.²⁴ While FOBT is a validated, cost-effective and non-invasive tool shown to reduce CRC-related mortality, its widespread use has also led to frequent misuse. Inappropriate use in patients who are not suitable candidates can result in unnecessary endoscopic procedures with low diagnostic yield and increased healthcare costs.¹⁴ However, recent studies have evaluated the role of FIT in symptomatic outpatients, prompting the British Society of Gastroenterology (BSG) and the Association of Coloproctology of Great Britain and Ireland (ACPGBI) to issue joint guidelines that better define the appropriate use of FIT in this population.

Clinical utility of FIT in symptomatic outpatients

In the United Kingdom, the 2-week wait (2WW) referral pathway was established in 2000 to expedite the evaluation and management of suspected CRC. Under this pathway, patients referred by a general practitioner (GP) for suspected cancer are expected to be evaluated by a specialist within 2 weeks and initiate treatment within 62 days.²⁵ To support timely and appropriate triage, the National Institute for Health and Care Excellence (NICE) has published guidelines recommending the use of FIT in primary care as a diagnostic tool for symptomatic patients.

NICE recommends offering FIT to adults with an abdominal mass, change in bowel habits or iron-deficiency anaemia as well as specific age and symptoms-based combinations. These include patients aged 40 and over with unexplained weight loss and abdominal pain; those under 50 with rectal bleeding and either abdominal pain or weight loss; individuals aged 50 and over with unexplained rectal bleeding, weight loss or abdominal pain; and adults aged 60 and over with anaemia, even in the absence of IDA.²⁶ These recommendations are designed to improve diagnostic efficiency by identifying high-risk individuals for urgent referral while reducing unnecessary investigations in those at lower risk.

Evidence supporting the diagnostic utility of FIT in this context is robust. A large prospective cohort study by Souza et al, known as the NICE FIT Study, evaluated the performance of FIT in 9822 symptomatic patients referred through the 2WW pathway across 50 hospitals in England.²⁷ Patients were stratified into high-risk and low-risk groups based on the NICE NG12 and DG30 guidelines, respectively. Using a threshold of 10 µg haemoglobin per gram of faeces (µg Hb/g), the study reported excellent diagnostic accuracy. In the high-risk group, FIT demonstrated a sensitivity of 92.2% (95% CI 88.2 to 95.2), specificity of 86.8% (95% CI 86.0 to 87.7), negative predictive value (NPV) of 99.6% (95% CI 99.4 to 99.8), and a positivity rate of 19.0%. In the low-risk group, sensitivity was 86.8% (95% CI 74.7 to 94.5), specificity 88.7% (95% CI 87.1 to 90.1), NPV 99.6 (95% CI 99.1 to 99.9) and positivity rate was also 19%. Importantly, only 0.3% of patients with a negative FIT result were ultimately diagnosed with CRC, confirming FIT’s value as an effective rule-out test. Overall, FIT correctly identified 96.6% of CRC cases in the high-risk group and 89.3% in the low-risk group, highlighting its utility across a broad spectrum of clinical presentations.

In response to this growing body of evidence, the BSG and the ACPGBI published joint guidelines to further define the role of FIT in evaluating symptomatic patients in the outpatient setting.²⁸ These guidelines reinforce the role of FIT in primary care as a triaging tool in patients with lower GI symptoms, recommending that a faecal haemoglobin concentration >10 µg Hb/g should prompt urgent referral for further investigation under the 2WW pathway. They also emphasise that patients presenting with an abdominal mass should be referred urgently, with FIT ordered simultaneously to inform downstream management. In cases of persistent anorectal bleeding, a negative FIT result should prompt further investigation with flexible sigmoidoscopy. Importantly, FIT is not recommended as a stand-alone exclusion tool; individuals with persistent, unexplained symptoms should be referred regardless of FIT results.

These guidelines provide a comprehensive and evidence-based framework for integrating FIT into CRC diagnostic pathways, ensuring high-risk patients receive timely care while minimising unnecessary procedures for those at low risk. Early data suggest that the widespread implementation of FIT in symptomatic outpatients is already reshaping referral patterns across the UK. GPs increasingly rely on FIT to stratify risk, leading to more targeted 2WW referrals and better use of limited endoscopic resources. Importantly, FIT helps reduce diagnostic delays for those most likely to benefit, while sparing low-risk patients from unnecessary colonoscopy and its associated burden.

Use and misutilisation of FOBT in the inpatient setting

In the inpatient setting, several factors can influence gFOBT results, increasing the risk of false positive or false negative and further reducing its clinical utility. These factors include patient age, comorbidities, dietary restrictions and medication use. Older adults—who represent a significant portion of hospitalised patients—are more likely to have GI conditions such as diverticulosis or peptic ulcers, which can result in false-positive FOBT results.²⁹ Additionally, medications commonly prescribed in this population, such as anticoagulants or NSAIDs, are known to cause false-positive outcomes in gFOBT. Bedridden or critically ill patients may also experience reduced gut motility, leading to constipation or minor GI trauma, which can further interfere with test accuracy. These complexities highlight the limitations of FOBT in accurately identifying GI pathology in the inpatient setting. While FOBT is primarily indicated for CRC screening, multiple studies have documented its frequent and often inappropriate use in the acute inpatient setting.3,1030 Unfortunately, many of these uses do not align with established protocols or indications.

In a retrospective study by Ip et al, none of the hospitalised patients undergoing FOBT received guidance regarding dietary or medication restrictions before testing. Among those who tested positive, 33% were taking acetylsalicylic acid, 16% were taking warfarin, 9% clopidogrel and 6% were NSAIDs—with some patients taking more than one of these medications at the time.⁸ Similarly, Friedman et al found that only 2% of the sample analysed had recommended dietary restriction before guaiac testing. In that study, 8% of patients had likely false-positive results due to inadequate dietary restrictions, all of whom later tested negative with immunochemical testing. In other instances, FOBT was inappropriately ordered for patients with clear evidence of overt GI bleed.⁷

These findings were similar to a study by Narula et al, which found that only 8.7% of patients undergoing FOBT in the inpatient setting had received medication restrictions. Among those on medications associated with false positives, 42.4% were on one such medication, 18.8% on two, and 23.1% on three—contributing to a 64% positivity rate on one or both tests. Notably, 11% of patients had normal GI endoscopic findings, further supporting the likelihood of false-positive results. Additionally, FOBT demonstrated low sensitivity in detecting true GI bleeding, especially in patients with iron-deficiency anaemia (IDA), as 42% of those with identifiable causes of IDA had false-negative results. This highlights the presence of false negatives and the test’s limited utility in patients with a high pretest probability of GI bleeding.⁶ This is further supported by a systematic review and meta-analysis by Lee et al, which reported a sensitivity of 58% for detecting endoscopically identifiable causes of IDA.³¹ This finding suggests that up to 42% of patients with treatable lesions may experience delays in care if endoscopy is deferred based on negative FOBT results.³¹

Moreover, studies have shown that many patients with suspected GI bleeding do not undergo a DRE as part of their initial evaluation. The DRE is a critical component of the physical exam in these cases, as it helps differentiate melena from bright red blood, guiding appropriate endoscopic evaluation. In a cohort study by Narula et al, only 18 patients (7.9%) who underwent FOBT had a documented DRE, with positive results in 7 (3.1%). Among these seven patients with a positive DRE, six had positive FOBT, but all seven were referred for an endoscopic investigation that identified a source of bleeding, suggesting that FOBT was likely unnecessary. Additionally, performing FOBT on samples collected during a DRE is discouraged, as anorectal trauma can result in false-positive results.⁶ Despite this, Soin et al found that 105 patients (14.4%) who underwent FOBT in the hospital had stool samples collected during a DRE, further highlighting the inappropriate use of the test.⁴

Beyond these specific examples, studies have identified broader patterns in the inpatient use of FOBT, particularly when used for diagnostic purposes. The most frequent indications for ordering FOBT in hospitalised patients included anaemia, followed by melena, abdominal pain and lower GI bleeding. Specialities most often responsible for these orders included, but were not limited to, Emergency Medicine, General Medicine, Intensive Care Unit and surgical services (table 3).46,9

Table 3. Common indications for ordering FOBT and involved specialties in the inpatient setting.

Study,year	Indications for ordering FOBT (n)	Amount of FOBT ordered by specialty (n)
Soin et al4	Anaemia (280)Melena (81)Abdominal pain (63)Lower GI bleed (112)Coffee ground emesis (17)Others (176)	Emergency department (not specified)General medicine Floor (not specified)Intensive care unit (not specified)
Friedman et al7	Anaemia for investigation (34)Symptoms consistent with GI bleeding (187)Non-bloody diarrhoea (15)Overt GI blood loss (17)Iron deficiency with or without anaemia (25)Screening for CRC (9)Pre-initiation of anticoagulation with warfarin (1)	General medicine (186)Specialty medical Units (82)Surgical specialties (43)Other specialities (19)
Ip et al8	Anaemia (270)Black stools (68)Overt gastrointestinal bleeding (32)Upper GI bleeding (16)Rectal bleeding (16)Gastrointestinal symptoms (51)Non-bloody diarrhoea (29)Abdominal pain or distention (8)Weight loss or weakness (4)Nausea or vomiting (4)Dysphagia (2)Bloody diarrhoea (4)Iron deficiency anaemia (3)Colorectal cancer screening (2)Before initiating anticoagulation (2)	Family medicine (158)Internal medicine (62)Orthopaedics (29)Intensive care (19)Psychiatry (13)General surgery (11)Other specialities (34)
Narula et al6	Anaemia (98)Overt bleeding (44)Suspected bleeding (63)Non-bloody diarrhoea (41)Iron deficiency (33)Dyspepsia (4)Colon cancer screening (1)Not documented (9)	General internal medicine (83)Gastroenterology (2)Other specialities of medicine, paediatrics or surgery (114)
Mosadeghi et al9	GI bleed (52)Anaemia (73)GI bleed+Anaemia (8)Unknown (68)	Not reported

CRC, colorectal cancer; FOBT, faecal occult blood test.

Clinical consequences of FOBT misutilisation

A positive FOBT in the inpatient setting often prompts a cascade of diagnostic tests and procedures aimed at identifying the underlying pathology. Common follow-up studies include esophagogastroduodenoscopy (EGD), capsule endoscopy, colonoscopy and CT. Patients with positive FOBT results are more likely to undergo endoscopic procedures than those with negative results.⁸ However, this approach carries several drawbacks. Care can be delayed while waiting for FOBT results, as providers may postpone specialist consultation, leading to increased length of stay and increased inpatient costs.⁶ Conversely, reliance on negative FOBT results can also delay care, even in cases where GI bleeding is clinically evident, as reported in a case by Saroja et al.³⁰

Available evidence suggests that FOBT use in the inpatient setting does not significantly improve patient outcomes. The diagnostic yield of performing endoscopic procedures following a positive inpatient FOBT is generally low. For instance, one study reported that only one CRC diagnosis was made for every 214 patients who underwent FOBT inappropriately.⁴ In a study of 5028 patients with positive FOBT results, half of the subsequent diagnostic procedures revealed no abnormalities, and only 17 new digestive malignancies were identified during follow-up.⁵

The financial burden associated with the workup prompted by positive FOBT results is another important consideration. Gupta et al estimated that these tests incurred a cost of approximately US $40 000 annually over 4 years.³ In addition to the economic impact, endoscopic procedures carry inherent risks, to which patients are exposed regardless of whether the procedure is ultimately warranted. Colonoscopies and EGD are associated with potential complications, including cardiopulmonary events related to sedation, haemorrhage, perforation, infection and, in rare cases, death (0.03%).^{32 33}

Educational initiatives and policy recommendations

Educating healthcare providers on the appropriate use of FOBT is essential for improving patient outcomes and ensuring diagnostic accuracy. Proper education reduces the risk of misdiagnosis, reduces unnecessary procedures and supports effective CRC screening. System-level solutions have been suggested as a method to address its misuse.³⁴ Educational initiatives that have shown to be effective include formal interdisciplinary meetings involving departments such as emergency medicine, hospitalists, gastroenterologists and the hospital’s Quality Council to review FOBT indications and performance characteristics.

Additionally, studies suggest that implementing electronic medical record (EMR) reminders, such as Best Practice Advisories, can significantly reduce unnecessary test utilisation in the emergency department and inpatient services. For example, in a study by Ajumobi et al, FOBT orders in the emergency department decreased from 3.10% of all visits to 0.09% after these implementations were integrated.¹⁰ Moreover, due to the remarkable decline in gFOBT use in the emergency department, all inpatient orders for gFOBT and FIT were removed from the hospital’s EMR order sets.

Raising awareness of the ongoing misuse of FOBT strengthens current practices and inspires new initiatives to improve test utilisation and patient safety across healthcare settings, creating a culture of evidence-based practice that ultimately improves the quality of care.

Conclusion

The inappropriate use of FOBT in the inpatient setting is a significant issue with far-reaching implications for patient care, resource utilisation and healthcare costs. Designed as a screening tool for asymptomatic, average-risk CRC patients in outpatient settings, FOBT is often misapplied in hospitals to evaluate anaemia, iron deficiency or GI bleeding. This misuse rarely impacts clinical decisions, frequently results in false-positive results due to impractical dietary or medication modifications and leads to unnecessary follow-up procedures, delays in diagnosis and prolonged hospital stays. To improve patient care and resource utilisation, healthcare providers must adhere to evidence-based guidelines, relying on thorough clinical assessments rather than routine FOBT in the acute setting. At the same time, strong evidence supports the use of FIT in symptomatic outpatients, where it aids in risk stratification and prioritisation for colonoscopy. Incorporating FIT appropriately in primary care and outpatient evaluations can streamline referrals, enhance early CRC detection and reduce the burden of unnecessary investigations—highlighting the importance of aligning test use with clinical context.

Quality improvement initiatives and interdisciplinary education for healthcare providers are essential to ensuring the proper application of FOBT and reducing its inappropriate utilisation. By fostering collaboration among clinicians, reinforcing evidence-based practices and discouraging FOBT misuse, healthcare systems can improve diagnostic accuracy, minimise costs and unnecessary interventions and enhance the overall quality and efficiency of inpatient care.

Data availability statement

Data sharing not applicable as no datasets generated and/or analysed for this study.