Using patient management as a surrogate for patient health outcomes in diagnostic test evaluation.

BACKGROUND: Before a new test is introduced in clinical practice, evidence is needed to demonstrate that its use will lead to improvements in patient health outcomes. Studies reporting test accuracy may not be sufficient, and clinical trials of tests that measure patient health outcomes are rarely feasible. Therefore, the consequences of testing on patient management are often investigated as an intermediate step in the pathway. There is a lack of guidance on the interpretation of this evidence, and patient management studies often neglect a discussion of the limitations of measuring patient management as a surrogate for health outcomes. METHODS: We discuss the rationale for measuring patient management, describe the common study designs and provide guidance about how this evidence should be reported. RESULTS: Interpretation of patient management studies relies on the condition that patient management is a valid surrogate for downstream patient benefits. This condition presupposes two critical assumptions: the test improves diagnostic accuracy; and the measured changes in patient management improve patient health outcomes. The validity of this evidence depends on the certainty around these critical assumptions and the ability of the study design to minimise bias. Three common designs are test RCTs that measure patient management as a primary endpoint, diagnostic before-after studies that compare planned patient management before and after testing, and accuracy studies that are extended to report on the actual treatment or further tests received following a positive and negative test result. CONCLUSIONS: Patient management can be measured as a surrogate outcome for test evaluation if its limitations are recognised. The potential consequences of a positive and negative test result on patient management should be pre-specified and the potential patient benefits of these management changes clearly stated. Randomised comparisons will provide higher quality evidence about differences in patient management using the new test than observational studies. Regardless of the study design used, the critical assumption that patient management is a valid surrogate for downstream patient benefits or harms must be discussed in these studies.

Prospective study of cardiac troponin I release in patients with upper gastrointestinal bleeding.

BACKGROUND AND AIM: The rate of cardiac injury in upper gastrointestinal hemorrhage is unclear. The aims of this study were to determine prospectively the risk of cardiac troponin I release and associated adverse cardiac events in patients with acute upper gastrointestinal hemorrhage. METHODS: From January to September 2003, we prospectively studied patients with documented hematemesis and melena referred to the gastroenterology unit in a tertiary teaching hospital in Melbourne, Australia. Serial assays for cardiac troponin I were performed at 0, 12 and 24 h. Serial creatine kinase levels and electrocardiographs were also performed. Clinical and biochemical data were collected. The primary endpoint was a troponin level >0.5 microg/L within 24 h of recruitment. Various clinical variables were then compared between the groups of patients with or without troponin rise. RESULTS: A total of 156 patients were included in the study. The mean age was 67 years (range 19-96). There were 104 (67%) male patients. A troponin level of greater than 0.5 microg/L was found in 30/156 (19%); 126 (81%) patients had normal troponin levels. Age greater than 65 years, signs of hemodynamic instability at presentation, a recent history of cardiac disease, cardiovascular compromise following endoscopy, and re-bleeding were associated with troponin release. CONCLUSION: Upper gastrointestinal bleeding is associated with a risk of cardiac injury of up to 19%. Troponin assay could be used to screen for cardiac damage, especially in elderly patients who present with hemodynamic instability.

Lamivudine resistance in patients with chronic hepatitis B: role of clinical and virological factors.

BACKGROUND: Lamivudine resistance is associated with long-term monotherapy for chronic hepatitis B and can lead to potentially serious clinical consequences. Scant information exists regarding the influence of hepatitis B virus variants in the development of resistance. The present study was designed to identify factors predictive of lamivudine resistance, with a particular focus on the role of precore and basal core promoter variants in the setting of hepatitis B e antigen-negative disease. METHODS: Eighty-five patients, representing four major genotypes, were followed prospectively on lamivudine therapy. Resistance was defined as an increase in viral load, with polymerase gene sequencing confirming a lamivudine resistance mutation. Median follow up was 19 months (6-54 months). The Cox proportional hazards model was used to determine variables independently predicting for the early onset of lamivudine resistance. RESULTS: The rate of lamivudine resistance was 6%, 31% and 51% at 12, 24 and 48 months, respectively. Multivariate analysis identified the precore variant, high baseline alanine aminotransferase (ALT), and persistent viremia (at 6 months) as independent predictors of the early development of lamivudine resistance, with rate ratios of 4.93 (95% confidence interval [CI]: 1.32-18.5), 1.22 (95%CI: 1.08-1.49), and 4.73 (95%CI: 1.49-15.0), respectively (P < 0.05). Female sex predicted early resistance (rate ratio 5.27 [95%CI: 1.23-22.5, P < 0.05]) although numbers were small (n = 12). Genotype did not influence treatment response nor time to onset of resistance. CONCLUSION: Patients with precore variant hepatitis B virus are likely to develop lamivudine resistance early and should be considered for alternate first-line monotherapy. In the future, combination antiviral therapy may limit the development of resistance.

Effect of the G1896A precore mutation on drug sensitivity and replication yield of lamivudine-resistant HBV in vitro.

Hepatitis B e antigen (HBeAg) negative chronic hepatitis B (CHB) is frequently caused by a mutation (G1896A) in the hepatitis B virus (HBV) precore (PC) reading frame that creates a stop codon, causing premature termination of the PC protein. During lamivudine treatment, drug resistance develops at a similar rate in HBeAg positive and HBeAg negative CHB. Lamivudine-resistant HBV mutants have been shown to replicate inefficiently in vitro in the absence of PC mutations, but it is unknown whether the presence of PC mutations affects replication efficiency or antiviral sensitivity. This study utilized the recombinant HBV baculovirus system to address these issues. HBV baculoviruses encoding the G1896A PC stop codon mutation were generated in wild-type (WT) and lamivudine-resistant (rtM204I and rtL180M + rtM204V) backgrounds, resulting in a panel of 6 related recombinant baculoviruses. In vitro assays were performed to compare the sensitivities of the PC mutant viruses with lamivudine and adefovir and to compare relative replication yields. The PC mutation did not significantly affect sensitivities to either adefovir or lamivudine. WT HBV and PC mutant HBV showed similar replication yields, whereas the replication yields of the lamivudine-resistant mutants were greatly reduced in HBeAg positive HBVs, confirming previous observations. However, the presence of the PC mutation was found to compensate for the replication deficiency in each of the lamivudine-resistant mutants, increasing the replication yields of each virus. In conclusion, the PC stop codon mutation appears to increase the replication efficacy of lamivudine-resistant virus but does not affect in vitro drug sensitivity.

Accuracy of EUS for detection of intraductal papillary mucinous tumor of the pancreas.

BACKGROUND: Patients with intraductal papillary mucinous tumors of the pancreas (IPMT) present with symptoms similar to those of chronic pancreatitis. This study assessed the accuracy of EUS for detection of IPMT and identified features that discriminate IPMT from chronic pancreatitis. METHODS: EUS accuracy for detecting IPMT was determined with characteristic findings by endoscopic retrograde pancreatography as the reference standard. To determine EUS features characteristic of IPMT, EUS images from patients with IPMT were compared with those from patients (similar age, gender) with chronic pancreatitis. RESULTS: Thirty-eight patients (23 men, 15 women; age range 40-90 years) with IPMT were identified between 1994 and 2001. For EUS, the sensitivity was 86%, specificity 99%, positive predictive value 78%, and negative predictive value 99% for detection of IPMT. When compared with patients with chronic pancreatitis, the EUS features of dilation of pancreatic duct (89% vs. 42%, p < 0.0001), cysts (45% vs. 11%, p = 0.002), and pancreatic atrophy (32% vs. 3%, p = 0.002) were more common, whereas parenchymal features of chronic pancreatitis were less common with IPMT (21% vs. 97%, p < 0.0001). By multivariate analysis, the presence of no more than one parenchymal feature of chronic pancreatitis suggested the diagnosis of IPMT (odds ratio 43.84; 95% CI [4.13, 465.74]). CONCLUSIONS: EUS may be useful in the initial evaluation of patients suspected to have IPMT. Paucity of parenchymal features of chronic pancreatitis is important in differentiating IPMT from other causes of chronic pancreatitis.

Push enteroscopy: a single centre experience and review of published series.

BACKGROUND: To assess the efficacy of push enteroscopy in a single tertiary hospital and review the available literature to assess the overall diagnostic yield of push enteroscopy. METHODS: Review of a database on push enteroscopy in a tertiary hospital from 1997 to 1999. This included 100 consecutive patients who underwent push enteroscopy. Review of all large published series on push enteroscopy to date to obtain an overall diagnostic yield. RESULTS: The diagnostic yield for patients with gastrointestinal (GI) bleeding was 47% and for patients with suspected small bowel disease was 33%. Angiodysplasia was the most common diagnosis in patients with GI blood loss. Patients with active GI bleeding had a higher diagnostic yield. The procedure was tolerated well and no complications occurred. Review of the literature showed an overall diagnostic yield of 44% (498 of 1 136 patients) for patients with GI blood loss and 38% (108 of 286 patients) for suspected small bowel disease. CONCLUSIONS: Push enteroscopy has a good diagnostic yield and is valuable in patients with GI blood loss and suspected small bowel disease.

Helical CT versus EUS with fine needle aspiration for celiac nodal assessment in patients with esophageal cancer.

BACKGROUND: Conventional CT is insensitive for detection of metastatic involvement of celiac lymph nodes in esophageal cancer. Helical CT has theoretical advantages over "slice" CT in this regard, but its performance has not yet been prospectively studied. METHODS: Consecutive patients with untreated esophageal cancer were recruited after obtaining informed consent. Helical CT was performed on all patients and TNM staging was performed by a single radiologist. Subsequently, all patients underwent esophageal radial and, as needed, curvilinear array EUS with fine needle aspiration (FNA), for evaluation of celiac lymph nodes and TNM staging. Test performance characteristics with 95% confidence intervals were calculated, assuming EUS with FNA as the reference standard. RESULTS: Forty-eight patients were recruited, of whom 37 (77%) were men. The mean (SD) age was 63.6 (10) years. Excluding 5 patients in whom a confirmatory FNA was not available (n = 43), helical CT identified celiac lymph nodes in 12 (28%) patients. The reference standard of EUS with FNA identified 15 (35%) patients with metastatic celiac lymph nodes, giving a sensitivity, specificity, and positive and negative predictive values for helical CT of 53% (95% CI [28%, 79%]), 86% (95% CI [73%, 99%]), 67% (95% CI [40%, 93%]), and 77% (95% CI [63%, 92%]), respectively, for assessing celiac lymph nodal involvement. The sensitivity and specificity of helical CT in detecting T4 disease were 25% (95% CI [3.8%, 46%]) and 94% (95% CI [85%, 100%]), respectively. There were 12 patients (25%; 95% CI [13%, 37%]) who were felt to have resectable disease by helical CT but had either metastatic involvement of celiac lymph nodes or T4 disease by EUS/FNA. CONCLUSIONS: Despite technological advances, helical CT still appears unreliable, mainly because of insensitivity, for the identification of inoperable T4 or metastatic involvement of celiac lymph node disease in esophageal cancer.