Mesenteric Venous Thrombosis Complicated by Heparin Induced Thrombocytopenia

Case Report with Review of Literature

Mesenteric Venous Thrombosis Complicated by Heparin Induced Thrombocytopenia

Corresponding author:  Dr. Gamal El Shallaly, Alzaiem Alazhari University, Ahmed Gasim Street, PO Box 2910, Khartoum North, Khartoum, Sudan, Tel: +249 907984078; Fax: +249 185 344510/12,
Email: gamalshallaly@hotmail.comgamalshallaly@rcsed.ac.uk

Abstract

Acute mesenteric venous thrombosis (MVT) is a rare cause of adynamic intestinal obstruction (paralytic ileus) that is associated with high mortality. A high index of suspicion is required to diagnose this condition. Contrast-enhanced CT scan is diagnostic. Management is usually complex and requires multidisciplinary approach, because MVT usually occurs in older patients with co-morbid disorders. Moreover, Heparin induced thrombocytopenia (HIT) can complicate heparin therapy which is the mainstay of medical management. Bowel ischaemia which may progress to infarction, gangrene and septicaemia is the main cause of morbidity and mortality. Hence, it is important to ensure intestinal viability, which may necessitate surgical resection that may be performed by laparotomy or laparoscopy. Additionally a second look laparotomy is sometimes indicated.

We present a case of MVT in a 67 year-old lady and review the published literature on this disorder. The lady was known diabetic hypertensive, as well as hepatitis C cirrhotic who also suffered from cardiac disease. We discuss the problems of diagnosis and management encountered in this patient within a clinical setting of a developing country. Despite a stormy postoperative course and intensive management, the patient survived with a good clinical outcome, and this has encouraged us to review the difficulties encountered in the diagnosis and management of this potentially fatal disorder.

Conclusion

MVT is a rare and fatal cause of small bowel obstruction. Its diagnosis requires a high level of suspicion. Early diagnosis and multidisciplinary management can largely improve its prognosis.

Keywords: Mesenteric venous thrombosis; Paralytic ileus; Heparin Induced Thrombocytopenia; Laparoscopy; Laparotomy; Second look Laparotomy; small bowel infarction

Introduction

Acute Mesenteric (bowel) Ischaemia (AMI) can be caused by either arterial or venous blockage. Arterial causes of AMI can  be due to superior mesenteric artery (SMA) thromboembolism and causes complete occlusion of the blood supply. It is, therefore, termed occlusive mesenteric ischaemia (OMI). SMA spasm/vasoconstriction causes non -occlusive ischaemia. SMA spasm is caused by low flow situations such as may occur in myocardial infarction, congestive heart failure, cardiac surgery, arrhythmias, shock, renal or hepatic disease Mesenteric venous thrombosis (MVT) results in non-occlusive mesenteric ischaemia (NOMI) and can lead to bowel infarction also.MVT is an uncommon cause of small bowel ischaemia that manifests with the clinical features of adynamic bowel obstruction (paralytic ileus). There are usually no specific symptoms and signs. This makes its early diagnosis difficult and causes high mortality due to intestinal infarction and gangrene.

MVT is now classified into acute, subacute, and chronic subtypes. Acute MVT is the most dangerous due to absence of collateralization of blood supply that occurs in the chronic and subacute types.

Acute MVT is defined as ‘new-onset symptomatic thrombosis of the superior mesenteric vein (SMV) or its branches without evidence of collateralization”[1]. The symptoms duration is less than 4 weeks. Non-specific abdominal pain is the most common symptom. Similar to arterial types of acute mesenteric ischemia (AMI), acute MVT may lead to intestinal infarction unless diagnosed and treated promptly. Subacute MVT and chronic MVT may be asymptomatic and some patients with chronic MVT present with portal hypertension-related bleeding oesophageal varices.

Incidence

The reported incidence of acute MVT varies widely in literature. It has been estimated that MVT cases account for 1 in 10000 emergency surgical admissions and less than 1 in 1000 laparotomies for acute abdomen[2,3]. It is notable, however, that the diagnosis of acute MVT has improved considerably, from being done at autopsy or laparotomy to CT scan diagnosis. This made possible the identification of acute MVT to be the cause of bowel ischaemia in (2.9-6%) of cases,[4,5] and in other estimates (5-15%)[6].

The rest are caused by arterial causes; SMA thrombo-embolism (60-70%), SMA spasm (20-30%)[6].

Regarding age, the median range is between 45-60 years, but it varies according to the aetiology of the venous thrombosis. For example, acute MVT caused by thrombophilias occur at an earlier age than those due to myeloproliferative diseases (MPD). There may be a slight male to female predominance.

There are no available data on the geographical pattern, although there has been an impression that MVT incidence could be higher in developing than developed countries due to the high prevalence of liver cirrhosis. However, there is no evidence to that and it is probably under diagnosed due to lack of emergency CT scanning and autopsy facilities in developing countries.

Pathophysiology

The main cause of morbidity and mortality in MVT is the development of intestinal ischaemia that can progress to infarction and gangrene leading to perforation, peritonitis, septicaemia and septic shock. The risk of bowel infarction depends on many factors including the site of the thrombosis, the rapidity of its formation and propagation, the development or not of collateral circulation, the extent (completeness) of vessel occlusion and interruption of blood flow.

It has been observed that in acute MVT, thrombosis of the smaller vessels (the intramural, vasa recta, and arcuate veins) causes more complete obstruction to the venous return and is, therefore, associated with higher risk of rapid bowel infarction and gangrene than thrombosis in the large splanchnic veins[7,8].

Hypercoagulability situations (see table) tend to cause thrombosis of the small veins and hence lead to early bowel infarction. In contrast, local factors tend to cause thrombosis in the large mesenteric and portal veins and hence the infarction occurs several days later[9,10].

The splanchnic veins mostly susceptible to thrombosis are the SMV, portal, and splenic veins. Thrombosis of the inferior mesenteric vein occurs rarely (4-7% of MVT). Hence, infarction occurs less commonly in the colon than the small bowel[8].

It has also been shown that infarction occurs more commonly with isolated SMV thrombosis than with combined SMV and portal vein or splenic vein involvement [7]. Isolated SMV thrombosis is rare, however, and is associated with a hypercoagulable state.

Mortality rate varies among various reports. It depends on whether or not bowel infarction has occurred. In case of infarction the mortality reaches (70%). The overall mortality rate in MVT approaches (44%) compared with (66-89%) of arterial thromboembolism[4]. Lower 30-day mortality rates of (10- 20%) could be achieved with early diagnosis and treatment, and the 2 year survival rate can reach (76%)[4,9,11].

Arterial vasospasm, though more common in arterial than venous occlusion is an important factor in the pathophysiology of ischaemia in MVT and may lead to bowel infarction after the start of anticoagulation therapy. Therefore, some authorities consider intra-arterial injection of papaverine at laparotomy to reverse the spasm[12].At the tissue and cellular level, in MVT the venous return is blocked and the bowel wall becomes engorged with blood and cyanotic. In contrast to cases with arterial occlusion, where there is a clear line of demarcation between normal and ischaemic bowel, in venous occlusion the transitional zone is wider and the change of colour is gradual[13]. This is seen at laparotomy and makes the decision on the level of bowel resection difficult to determine. This led some authorities to advocate doing a wider excision, [14] or do nothing and have a second look laparotomy 24-48 hours later in the hope that more demarcation occurs[15].

The epithelium is known to be affected by ischaemia earlier than the other layers of the bowel due to its high metabolic rate. Epithelial changes and necrosis can thus occur before transmural necrosis and changes of the colour of bowel wall occurs. It was shown that non operative treatment with anticoagulants can avoid the need to surgery as long as transmural necrosis has not occurred [11]. However, due to the risk of translocation of bacteria from bowel lumen into the blood stream the patient must be covered with intravenous broad spectrum antibiotics.

Laparotomy can hence be reserved for those patients with abdominal signs of peritonitis caused by bowel perforation secondary
to transmural infarction, or if there are any symptoms and signs suspicious of septicaemia even if abdominal signs are equivocal.

Bowel ischaemia results in paralytic ileus. Venous congestion leads to transmural oedema which gives the diagnostic concentric thickening of the bowel wall and halo sign on CT scan. Intraluminal haemorrhages may occur and leads to symptomatic bleeding per rectum.

Clinical Presentation

Acute MVT presents with vague non- specific symptoms and signs. High index of suspicion is required to diagnose it.

Symptoms

The most common symptom is abdominal pain, which could be diffuse, or localized mainly in the upper or lower quadrants[ 1,11]. Other symptoms include nausea and vomiting, abdominal distension and constipation. All these symptoms constitute adynamic bowel obstruction, or paralytic ileus. Diarrhea and bleeding per rectum were also described in some patients[1].

It is important to ask about past or family history of deep vein thrombosis, as this may indicate the presence of a hereditary thrombophilia.

Signs

On clinical examination, it is important to check the vital signs. Initially there may be low grade fever 37.7 (+/- 0.1)[11]. The presence of septic shock and/or guarding and rigidity on abdominal examination is indicative of bowel necrosis and is an indication of surgery.

Investigation

The most useful tool to diagnose MVT is IV contrast- enhanced CT scan[6,11]. Multidetector CT scans have two dimensional arrays of detector elements and higher resolution than conventional and helical CT scanners[15]. The most important CT diagnostic features of MVT associated bowel ischaemia are the presence thrombosis with concentric bowel wall thickening (>3mm) with hypoattenuation[16].

Plain abdominal x-ray shows non -specific features of bowel obstruction such as bowel dilatation (supine views) and air fluid levels (erect views). Radiological features such as thumb printing, pneumaosis intestinalis and portomeseneric venous gas indicate bowel gangrene in a high proportion of cases (>75%). They are more common in arterial than venous ischaemia, and are considered as ominous signs[17].

Table 1. Showing the development of Heparin induced thrombocytopenia (HIT) after starting heparin infusion.

The gold standard of diagnosis is splanchnic angiography. This has both diagnostic and therapeutic values as it may be used for intrarerial vasodilator therapy. However, due to their invasiveness and unavailability in most hospitals these modalities are reserved for the few clinically suspicious cases in which CT scan was equivocal[1]. Other potentially useful diagnosticmodalities are Magneic resonance angiography (MRA) and Doppler ultrasound.

Laboratory tests include full blood count which usually displays leukocytosis. Platelets are initially normal but should be monitored to detect the earliest sign of anticoagulant therapy complication termed heparin induced thrombocytopenia (HIT).

Table 2. Shows reported predisposing factors for MVT.

It is important to do coagulation screen, such as APTT and PT as baseline to anticoagulation therapy. Urea, creatinine, andelectrolytes reflect hydration and renal function. Hypokalaemia, if present, must be corrected as it causes or worsens paralyticileus. Arterial blood gases (ABGs), and liver function tests are important. Lactic acidosis reflects late ischaemia. Impaired liver function due to cirrhosis is an important cause of secondary MVT (Table 2).

Treatment

There is currently a consensus of opinion that the initial treatment of MVT is non- operative via anticoagulation[6,11,12].Exploratory laparotomy is indicated only if there is evidence (clinical, or radiological) of transmural bowel necrosis. Resection of macroscopically necrotic bowel is done followed by primary anastomosis or ileostomy. Sometimes a second look laparotomy is advocated since the transitional zone between normal and necrotic bowel is frequently not clear in the first laparotomy.

Anticoagulation and its complications

HIT is a known complication of heparin prophylaxis and therapy. It is caused by the antibody IgG forming a complex with platelet factor 4 (PF4) and heparin. The IgG/PF4/heparin complexes then bind to and activate platelets. Platelets are consumed causing thrombocytopenia. Paradoxically the complexes may regenerate thrombin thus creating prohrombotic conditions causing the formation of arterial and venous thrombi.

The incidence of HIT is (0.2%-5%) in those receiving heparin. The risk is greater in thromboprohylactic than therapeutic doses, with unfractionated heparin (UFH) than low molecular weight heparin (LMWH), and in surgical patients than medical ones. It is more common with bovine than porcine heparin[18].

The Haemostasis and Thrombosis Task Force of the British Committee for Standards in Haematology (BCSH) recommends that postoperative patients on heparin should have a baseline platelet counts before starting heparin, and every 2-3 days from day4-14 postop or until heparin is stopped[18].HIT should be considered clinically if four features are present in the patient. A 30% or more drop in the platelet, occurringbetween day 4 and 14 after heparin administration, in the absence of an obvious cause for thrombocytopenia, and with evidence of previous or new thrombosis. These 4 features form the basis of a pretest ‘4Ts’ scoring system. A low score excludes the diagnosis. A high score necessitates stopping heparin and giving alternative anticoagulant and performing laboratory confirmatory tests of HIT antibodies. These are classified into platelets activation assays and immunological assays[18]. These specific tests, however, are not readily available except in specialized labs, hence the diagnosis remains clinical in most situations.

Once HIT is diagnosed, heparin should stop and replaced by alternative anticoagulants. These are broadly classified into: (i)direct thrombin inhibitors such as dabigatran (ii) indirect activated factor ten (Xa) inhibitors such as Fondaparinux, which are like heparins in that they inhibit free factor Xa but not that bound to the prothrombinase complex, and are given subcutaneously. (iii) Direct Xa inhibitors such as Rivaroxaban, apixaban, and edoxaban, which have many properties of an idealanticoagulant such as oral administration, rapid onset of action, and predictable pharmacokinetics and pharmacodynamics.However, they lack specific antidote to reverse their action if bleeding occurs.

Case Presentation

A 62-year-old Saudi female patient presented to the emergency department (ED), complaining of abdominal pain and vomitingfor four days.

She was in her usual state of health untill 4 days prior to presentation when she noticed a sudden onset of abdominal pain. The pain was diffuse, and colicky with no radiation. It was aggravated by meal, and no relieving factor was identified. The pain was progressing with time, with severity approaching 9/10 on the visual analogue scale. The pain was associated with frequent vomiting of food content but no blood. She also noticed abdominal distention starting with the onset of pain. She had constipation for 3 days. There was no fever, no melena or blood in stools, and no weight loss.

Her past medical history included diabetes mellitus (DM), hypertension (HTN), liver cirrhosis due to hepatitis C virus (HCV), and cardiomyopathy. She had no past history of a similar attack or surgical operations. She had a positive family history (a brother) of upper limb deep venous thrombosis (DVT).

On clinical examination: She was conscious, but drowsy, and in pain. The vital sign were stable and she was afebrile. Therewere no stigmata of chronic liver disease. Chest examination was normal.

The abdomen looked distended, with full flanks, and flat umbilicus but no other abnormality.

On palpation, the abdomen was soft. There was some tenderness, mainly at the peri-umbilical and left lower quadrant areasbut no guarding. Percussion revealed a positive shifting dullness. Bowel sounds were absent. The rectum was empty and there was no blood on digital rectal examination (DRE).

Her laboratory investigations showed: WBC: 18 x109/L, HB: 12gm/dl, and platelet: 210 x109/L.

Arterial blood gases, with patient breathing room air were; ABG: PH: 7.4, HCO3 (24 mmol/L), PaCO2 (38mm Hg), and PaO2 (90 mmHg).

Abdominal X-ray showed: dilated small bowel with multiple air-fluid levels (Figure 1. a, b)

Figure 1(a). Supine plain abdominal x-ray showing dilated small bowel loops

Figure1(b). Erect plain abdominal x-ray showing multiple air-fluid levels

Liver function tests showed bilirubin 0.8 mg/dl, albumin 2.1, INR 1.2.

The patient was admitted with the provisional diagnosis of small bowel obstruction. CT abdomen with IV and oral contrast was done and showed subtotal portal and superior mesenteric venous thrombosis complicated with small bowel obstruction, and cirrhotic liver with mild ascites (Figure 2. a, b). The CT result was obtained on the 2nd day of admission

Figure 2(a). Contrast-enhanced CT scan showing portal and SMV thrombosis: Cross section

Figure 2(b). Contrast-enhanced CT scan showing portal and SMV thrombosis Longitudinal section.

A decision was then made to start intravenous therapeutic heparin, with a loading dose of 5000 Units followed by infusion at a rate of 1000 IU/hr. Her baseline platelet was 210, INR: 1.2 and activated partial thromboplastin time (aPTT) 37 seconds.

She was admitted in the intensive care unit (ICU)

She developed a temperature of (38.50 C) the following day and there was no improvement in her condition. She becameconfused and didn’t recognize her relatives. Septicaemia was suspected as a cause of the confusion and fever. CT scan of thebrain to exclude intracranial thrombosis was requested but it was not expected to be done and reported expeditiously enough. A decision was therefore taken to explore the abdomen. Heparin infusion was stopped and two hours later the patient was taken to the operating room (OR). Laparoscopic exploration was initially attempted but dilated small bowel loops obstructed the view. It was, hence, converted to a long midline laparotomy. An ischemic but not frankly gangrenous segment of the small bowel was identified. This was about 30cm long and 130cm from the duodenojejunal (DJ) junction. After 100% oxygenation and warming the bowel with warm saline, the viability was tested using color, peristalsis, and mesenteric pulsation. It showed improvement in the color, peristalsis and mesenteric pulses were good. A decision was made to close the abdomen leaving a drain, and a second look laparotomy planned after 48 hours. IV heparin started 12hours postoperatively for fear of oozing from the wound.

In the second look surgery, the bowel was viable and no resection was needed (Figure 3. a, b).

Figure. 3 (a). Before warm saline packing. Bowel is viable. No need for resection

Figure 3(b). Bowel after warm saline packing

Figure 3. showing the colour of the ischaemic bowel during the 2nd look laparotomy

The abdomen was closed again and IV heparin restarted 6 hours postoperatively, and titrated by aPTT monitoring, aiming at a level of 2-3 times the normal.

The patient was responding well to the treatment. She started oral fluids after the bowel sounds were heard and she passedstools after 3 days from the 2nd look operation. There was, however, a high output of ascetic fluid through the drain; about2000 ml / day. After starting orally she was transferred from the ICU to the high dependency unit (HDU)

While in the ICU, She dropped her platelets to more than 65% of the baseline count after four days from starting the heparininfusion (Table 1).

HIT was, therefore, suspected. Heparin was stopped and the patient was examined for any signs of thrombosis. Alternativeanticoagulants were considered. A multidisciplinary committee was formed from ICU consultant, a general physician -as no haematologist was available- and the surgical team. Fondaparinux was considered the most suitable alternative anticoagulant considering her liver condition. However, this drug was not available in our hospital. So we started her on Rivaroxaban. This has the convenience of oral once daily dosing without the need for monitoring. She continued to show improvement both clinically and in her platelets count. HIT confirmatory tests were not available and HIT diagnosis remained largely clinical.

The patient also received human albumin due to hypoalbuminemia.

She suffered from a mild wound infection that was treated by a combination of wound dressing and vacuum suction.

The patient was responding well to the management apart from an episode of sub-segmental pulmonary embolism (PE)despite the anticoagulation. This was more likely a manifestation of HIT, and was treated conservatively.

She showed thereafter progressive improvement clinically and no more ascetic fluid was obtained through the drain. Herplatelets count increased to 150 and she was discharged in good condition after 4 weeks of hospitalization. A CT scan of the brain was done and that was normal.

She was seen in the outpatient clinic for 6 weeks after discharge and she was doing well. She had no more abdominal pain and her platelets count was within normal limits. The plan was to continue on rivaroxaban for life and to be seen with the physicians.

Discussion

The patient who was 67 year old and suffering from several comorbidities presented with symptoms and signs of paralytic ileusof about 4 days before admission. The comorbidities were diabetes mellitus, hypertension, cardiomyopathy, and hepatitis C liver cirrhosis. She was Child’s-Pugh class B. These comorbidities did contribute significantly to her morbidity and thecomplexity of the management, and made a multidisciplinary approach mandatory.

The CT scan was diagnostic. It showed MVT affecting both the portal vein and SMV. Therapeutic intravenous (IV) unfractionated heparin (UFH) therapy was started upon CT diagnosis on the second day of admission. This delay didn’t seem to have affected the outcome adversely.

The treatment could have continued with this non-operative approach[11]. However, the development of confusion and fevermade us suspicious of septicaemia due to bowel infarction, and was our indication for early laparotomy. IV heparin wasstopped and two hours later the patient was taken to the OR.

Laparoscopy was attempted first but visibility was poor due to bowel distension and therefore a laparotomy was done through a long midline incision. This enabled full inspection of the bowel and identification of the affected segment which was judged to be still viable after covering the bowel with packs soaked in warm saline for about 5 minutes. We were able to palpate the mesentery and its pulses and squeeze the bowel and see if it had peristalsis. These were the advantages of laparotomy over laparoscopy in this case. A decision was then taken to close the abdomen without resecting the bowel, and to do a 2nd look laparotomy after 48 hours. This was also done via opening the recent scar and the bowel was found to be viable. A decision was taken to close and continue heparin therapy as it was obviously successful.

The aetiology of MVT in this patient seemed to be multifactorial; both local and systemic. The local predisposing factormentioned in the literature and found in this patient was liver cirrhosis. She also had a strong family history of DVT in abrother. Hence, some form of systemic hereditary hypercoagulability disorder may also be present. However, it seemed thatthe local factor was the predominant one, because thrombosis affected the large splanchnic vessels (both portal and SMV).Accordingly there was delayed progress of thrombosis and bowel infarction, so that the early anticoagulation therapy wassuccessful[15,16]. Systemic hypercoagulable states were observed to affect smaller vessels and lead to early bowel infarction [9,10]. Unfortunately we could not ascertain the presence or absence of a hereditary thrombophilia due to the lack of specific tests.

The question of when to stop and restart therapeutic heparin in a patient requiring anticoagulation and undergoing laparotomyis challenging. This is because interrupting anticoagulation increases the risk of thromboembolism, whereas continuing it increases the risk of bleeding. There are no randomized trials and the decision is mainly based on expert opinion and clinical judgement that strike a balance between both risks. Due to the short dose dependent elimination half- life of UFH (90 minutes), the American College of Chest Physicians Evidence-Based Clinical Practice Guidelines suggest stopping UFH 4-6 hours before surgery (and LMWH 24 hours before surgery)[19]. Some authorities even advocate starting UFH intra-operatively if MVT has not been diagnosed preoperatively[ 11]. Regarding postoperative resumption of therapeutic dose of UFH American College of Chest Physicians recommend that LMWH or UFH should not be resumed at a fixed time after a surgery or procedure without consideration of the anticipated bleeding risk or adequacy of postoperative hemostasis[19]. Some authorities give it immediately after the operation[11]. We have resumed therapeutic UFH with the same infusion rate (no bolus dose), after 12 hours after the first operation and 6 hours after the 2nd look operation without adverse effects and no bleeding occurred.

The patient was discharged eventually in a good condition. Her hospital stay was within the expected time as reported by other studies[11]. Since the aetiology is mainly cirrhosis she should continue on anticoagulant for life. Because future complicationssuch as bowel stenosis and secondary perforation may occur (<10% of cases), the patient should be followed up closely.

Conclusion

MVT is a rare but fatal cause of acute abdomen and paralytic ileus. It can have a good outcome if diagnosed and treated early. Anticoagulant is the mainstay of treatment but beware of heparin induced thrombocytopenia. The comorbidities and complexity of these cases make multidisciplinary approach essential. The surgeon should be prepared to do laparotomy and even 2nd look laparotomy if there is any suspicion of bowel infarction.

References

1.Harnik IG, Brandt LJ. Mesenteric Venous Thrombosis. Vasc Med. 2010, 15 (5): 407-418.

2.Hansen HJB, Christoffersen JK. Occlusive mesenteric inf¬arction: a retrospective study of 83 cases. Acta Chir Scand Suppl. 1976, 472: 103–108.

3.Kazmers A. Intestinal ischemia caused by venous throm¬bosis. In: Rutherford R (ed). Vascular surgery. 4th edition. Philadelphia: WB Saunders, 1995.

4.Corton MM, McIntire DD, Twickler DM, Atnip S, Schaffer JI et al. Endoanal ultrasound for detection of sphincter defects following childbirth. Int Urogynecol J. 2013, 24(4): 627-635.

5.Walter SD, Eliasziw M, Donner A. Sample size and optimal designs for reliability studies. Stat Med. 1998, 17(1):101-110.

6.Cappelleri JC, Ting N. A modified large-sample approach to approximate interval estimation for a particular intraclass correlation coefficient. Stat Med. 2003, 22(11):1861-1877.

7.Fleiss JL, Nee JC, Landis JR. Large sample variance of kappa in the case of different sets of raters. Psychol Bull. 1979, 86(5):974-977.

8.Landis JR, Koch GG. The Measurement of Observer Agreement for Categorical Data. Biometrics. 1977, 33(1):159-174.

9.Woehr DJ, Huffcutt AI. Rater training for performance appraisal: A quantitative review. J Occup Organ Psychol. 1994, 67(3): 189-205.

10.Gold DM, Halligan S, Kmiot WA, Bartram CI. Intraobserver and interobserver agreement in anal endosonography. Br J Surg 1999, 86(3): 371-375.

11.Fowler GE, Adams EJ, Bolderson J, Hosker G, Lowe D et al. Liverpool Ultrasound Pictorial Chart: the development of a new method of documenting anal sphincter injury diagnosed by endoanal ultrasound. BJOG. 2008, 115(6): 767-772.

12.Abdool Z, Sultan AH, Thakar R. Ultrasound imaging of the anal sphincter complex: a review. Br J Radiol. 2012, 85: 865- 875.

13.Richter HE, Nager CW, Burgio KL. Incidence and Predictors of Anal Incontinence After Obstetric Anal Sphincter Injury in Primiparous Women. Female Pelvic Med Reconstr Surg. 2015, 21(4): 182-189.

14.Corton MM, McIntire DD, Twickler DM, Atnip S, Schaffer JI et al. Endoanal ultrasound for detection of sphincter defects following childbirth. Int Urogynecol J. 2013, 24(4): 627-635.

15.Walter SD, Eliasziw M, Donner A. Sample size and optimal designs for reliability studies. Stat Med. 1998, 17(1):101-110.

16.Cappelleri JC, Ting N. A modified large-sample approach to approximate interval estimation for a particular intraclass correlation coefficient. Stat Med. 2003, 22(11):1861-1877.

17.Fleiss JL, Nee JC, Landis JR. Large sample variance of kappa in the case of different sets of raters. Psychol Bull. 1979, 86(5):974-977.

18.Landis JR, Koch GG. The Measurement of Observer Agreement for Categorical Data. Biometrics. 1977, 33(1):159-174.

19.Woehr DJ, Huffcutt AI. Rater training for performance appraisal: A quantitative review. J Occup Organ Psychol. 1994, 67(3): 189-205.

Be the first to comment on "Mesenteric Venous Thrombosis Complicated by Heparin Induced Thrombocytopenia"

Leave a comment

Your email address will not be published.


*