Ten common statistical errors and how to avoid them

Impact of patient age on outcome after resection for hepatocellular carcinoma

There is little information on the impact of aging on liver resection of hepatocellular carcinoma (HCC). The aim of study was to evaluate the prognostic impact of the patient's age on the long-term survival after resection of HCC. The postoperative outcomes of the 291 elderly (≥ 70 years) and 340 younger (< 70 years) patients underwent curative liver resection for HCC were analyzed using multivariate and propensity-score matching. Risk score were calculated from the results of Cox regression analysis. The overall survival rate was significantly lower in the elderly group than that in the younger group (p = 0.01). Factors related to overall survival were vascular invasion (absent vs. present, HR 2.25; 95% CI 1.52-3.33, p = 0.0001), albumin level (< 3.0 vs. ≥ 3.0 g/dl, HR 2.23; 95% CI 1.31-3.79, p = 0.003), and number of tumors (solitary vs. multiple, HR 1.68; 95% CI 1.24-2.27, p = 0.001). The results of risk-score analysis with a Cox proportional-hazards model indicated that the proportion of poor-risk patients was significantly higher in the elderly than in the younger group. Propensity-score matching analysis yielded 234 pairs of patients. There were no significant differences in baseline profiles or risk scores between the two groups (p = 0.43). There were also no significant differences in the overall survival between the two groups (p = 0.23). Advanced age does not have a significant impact on the outcomes of patients after resection of HCC.

Laparoscopic Liver Resection versus Percutaneous Radiofrequency Ablation for Small Single Nodular Hepatocellular Carcinoma: Comparison of Treatment Outcomes

BACKGROUND

Treatment outcomes of laparoscopic liver resection (LLR) and percutaneous radiofrequency ablation (p-RFA) for small single hepatocellular carcinomas (HCCs) have not yet been fully compared. The aim of this study was to compare LLR and p-RFA as first-line treatment options in patients with single nodular HCCs ≤3 cm.

METHODS

From January 2014 to December 2016, a total of 566 patients with single nodular HCC ≤3 cm treated by either LLR (n = 251) or p-RFA (n = 315) were included. The recurrence-free survival (RFS) and cumulative incidence of local tumor progression (LTP) were estimated using Kaplan-Meier methods and compared using the log-rank test. Treatment outcome of 2 treatment modalities was compared in the subgroup of patients according to the tumor location.

RESULTS

There were no significant differences in overall survival between LLR and p-RFA (p = 0.160); however, 3-year RFS was demonstrated to be significantly higher after LLR (74.4%) than after p-RFA (66.0%) (p = 0.013), owing to its significantly lower cumulative incidence of LTP (2.1% at 3 years after LLR vs. 10.0% after p-RFA, p < 0.001). The complication rate of p-RFA was significantly lower than that of LLR (5.1 vs. 10.0%, p = 0.026). LLR also provided significantly better local tumor control than p-RFA for subscapular tumors (3-year LTP rates: 1.9 vs. 8.8%, p = 0.012), perivascular tumors (3-year LTP rates: 0.0 vs. 17.2%, p = 0.007), and tumors located in anteroinfero-lateral liver portions (3-year LTP rates: 0.0 vs. 10.7%, p < 0.001). However, there were no significant differences in LTP rates between LLR and p-RFA for non-subcapsular and non-perivascular tumors (p = 0.482) and for tumors in postero-superior liver portions (p = 0.380).

CONCLUSIONS

LLR can provide significantly better local tumor control than p-RFA for small single HCCs in subcapsular, perivascular, and anteroinferolateral liver portions and thus may be the preferred treatment option for these tumors.

Impact of marginal resection for hepatocellular carcinoma

PURPOSE

The surgical margin for liver resection to treat hepatocellular carcinoma (HCC) is occasionally < 1 mm. This study determined the impact of a surgical margin < 1 mm [marginal resection (MR)] on the types of recurrence and the prognosis in solitary HCC.

METHODS

The data of 454 patients undergoing curative liver resection for solitary HCC in our institution were analyzed. The patients were divided into the MR (n = 90) and non-MR (n = 364) groups. The clinicopathological data and outcomes after liver resection were compared. A case-matching analysis using a propensity scoring method was also performed.

RESULTS

The recurrence-free survival was significantly and overall survival was marginally significantly lower in the MR group than in the non-MR group (p = 0.012-0.051, respectively). According to a multivariate analysis, MR was not a significant independent factor for recurrence-free survival (p = 0.056). After propensity score matching, there were no significant differences in the recurrence-free and overall survival between the two groups (p = 0.375-0.496, respectively). Furthermore, there were no significant differences in the intrahepatic recurrence patterns between the two groups before and after matching.

CONCLUSION

MR for solitary HCC might be sufficient in patients with a limited liver functional reserve.

Ginsenoside-Rb1 for Ischemic Stroke: A Systematic Review and Meta-analysis of Preclinical Evidence and Possible Mechanisms

Background

Ischemic stroke is the most common type of stroke, while pharmacological therapy options are limited. Ginsenosides are the major bioactive compounds in Ginseng and have been found to have various pharmacological effects in the nervous system. In the present study, we sought to evaluate the effects of Ginsenoside-Rb1 (G-Rb1), an important ingredient of ginsenosides, and the probable neuroprotective mechanisms in experimental ischemic strokes.

Methods

Studies of G-Rb1 on ischemic stroke animal models were identified from 7 databases. No clinical trials were included in the analysis. The primary outcome measures were neurological function scores, infarct volume, evans blue content and/or brain water content (BWC). The second outcome measures were the possible neuroprotective mechanisms. All the data were analyzed by Rev Man 5.3.

Result

Pooled preclinical data showed that compared with the controls, G-Rb1 could improve neurological function (Zea Longa (n = 367, P < 0.01); mNSS (n = 70, P < 0.01); Water maze test (n = 48, P < 0.01); Bederson (n = 16, P < 0.01)), infarct area (TTC (n = 211, P < 0.01); HE (n = 26, P < 0.01)), as well as blood-brain barrier function (BWC (n = 64, P < 0.01); Evans blue content (n=26, P < 0.05)). It also can increase BDNF (n = 26, P < 0.01), Gap-43 (n = 16, P < 0.01), SOD (n = 30, P < 0.01), GSH (n = 16, P < 0.01), Nissl-positive cells (n = 12, P < 0.01), Nestin-positive cells (n = 10, P < 0.05), and reduce Caspase-3 (n = 36, P < 0.01), IL-1 (n = 32, P < 0.01), TNF-α (n = 72, P < 0.01), MDA (n = 18, P < 0.01), NO (n = 44, P < 0.01), NOX (n = 32, P < 0.05), ROS (n = 6, P < 0.05), NF-κB (P < 0.05) and TUNEL-positive cells (n = 52, P < 0.01).

Conclusion

Available findings demonstrated the preclinical evidence that G-Rb1 has a potential neuroprotective effect, largely through attenuating brain water content, promoting the bioactivities of neurogenesis, anti-apoptosis, anti-oxidative, anti-inflammatory, energy supplement and cerebral circulation.

Statistical Accuracy in Rheumatology Research

Science and related research activities are always subject to errors. Statistical accuracy is necessary in order to overcome erroneous researching and reporting practices. The present article aimed to review the current literature on statistical errors in clinical medicine articles, and to provide rheumatologists with basic recommendations regarding the use of common statistical methods in research articles. With this purpose, PubMed/MEDLINE and Web of Science databases were searched by using relevant keywords. Data so far indicate that statistical errors are common in published articles from several disciplines of medicine. Statistics is the key element of any research activity, thus, implementing statistics at each step (hypothesis development, study design, sampling/data collection, data analysis, presentation) of every research is mandatory. In this regard, awareness of common statistical errors, basic knowledge on statistical methodology and consulting an expert in biostatistics from the beginning of the research process would be of value for rheumatologists.

Postoperative Portal Hypertension Enhances Alloimmune Responses after Living-Donor Liver Transplantation in Patients and in a Mouse Model

Controlling portal vein pressure in living-donor liver transplantation has received increased attention owing to its potential importance for graft survival. Portal hypertension may lead to the activation of liver-resident APCs, including liver sinusoidal endothelial cells (LSECs), which have immunological tolerogenic capacity. We investigated the effects of portal hypertension on graft survival and the antidonor immune response using clinical data and a mouse model. We categorized patients (n = 136) according to their portal vein pressure values at the end of surgery. Using propensity score-matching analyses, we found that portal hypertension was significantly associated with a higher antidonor immune response and incidence of acute rejection. To investigate the mechanism, we performed an allogeneic coculture assay using a 70% hepatectomized (HTx) mouse model with or without a portosystemic shunt. Liver cells from HTx mice without a shunt exhibited a significantly greater anti-BALB/c B6 T cell response than those from sham-operated mice or HTx mice with a shunt. LSECs from sham-operated mice, but not from HTx mice, suppressed the B6 T cell alloresponse in a dose-dependent manner. Furthermore, LSECs from HTx mice without a shunt showed significantly downregulated MHC class I/II and programmed death-ligand 1 expression, and those from mice with a shunt showed recovered expression of these molecules. Postoperative portal hypertension enhances alloimmune responses in recipients after living-donor liver transplantation, likely due, in part, to the impaired immune-suppression capacity of LSECs.

Comparison of hepatic arterial infusion chemotherapy between 5-fluorouracil-based continuous infusion chemotherapy and low-dose cisplatin monotherapy for advanced hepatocellular carcinoma

AIM

The aim of this study is to compare the efficacy and safety of hepatic arterial infusion chemotherapy (HAIC) between 5-fluorouracil (5-FU)-based continuous infusion chemotherapy and low-dose cisplatin (CDDP) monotherapy in patients with advanced hepatocellular carcinoma (HCC).

METHODS

Patients were grouped according to HAIC regimen (5-FU group, n = 317/CDDP group, n = 66). A two-to-one match was created using propensity score analysis (5-FU group, n = 102/CDDP group, n = 51). After matching, response rate (RR) and adverse events as primary end-points, and survival and progression-free survival as secondary end-points, were analyzed.

RESULTS

In the analysis of primary end-points, the RR in the 5-FU group was significantly higher than in the CDDP group (32.4% vs. 15.7%, P = 0.033). In patients with a Child-Pugh (CP) score of 5-7, the RR in the 5-FU group was significantly higher than that in the CDDP group (36.1% vs. 15.4%, P = 0.020). In those with a CP score of 8-9, there was no significant difference in RR between the two groups (15.8% vs. 16.6%, P = 1.000). The reservoir system-related complications were 9.8% in the 5-FU group, and there was no significant difference in the incidence of grade 3/4 adverse events between the two matched groups (P > 0.05). In terms of secondary end-points, the median survival time was 9.1 and 8.7 months for the 5-FU and CDDP groups, respectively (P = 0.4917). Progression-free survival was 3.9 months for the 5-FU group and 4.9 months for the CDDP group (P = 0.4).

CONCLUSIONS

5-Fluorouracil-based continuous infusion chemotherapy could be suitable for advanced HCC patients with a CP score of 5-7 considering the treatment response.

The risks of hepatocellular carcinoma development after HCV eradication are similar between patients treated with peg-interferon plus ribavirin and direct-acting antiviral therapy

The risk of hepatocellular carcinoma (HCC) development is reduced following viral elimination by interferon therapy in chronic hepatitis C patients. However, the risk in patients treated with interferon-free direct-acting antivirals (DAAs) is unknown. We evaluated chronic hepatitis C patients who achieved viral eradication by pegylated-interferon plus ribavirin (PEG-IFN/RBV, n = 244) or daclatasvir plus asunaprevir (DCV/ASV, n = 154) therapy. None of the patients had prior history of HCC or antiviral therapy. The median observation period after the end of treatment for the PEG-IFN/RBV and DCV/ASV groups were 96 (range 10–196) and 23 (range 4–78) months, respectively. During the observation period, HCC developed in 13 (5.3%) and 7 (4.5%) patients in the PEG-IFN/RBV and DCV/ASV groups, respectively. The cumulative HCC development rate after 1-, 3- and 5-years (0.4%, 3% and 5% for the PEG-IFN/RBV group and 0.6%, 9% and 9% for the DAA group, respectively) were similar between the two groups. Propensity score matching analysis also showed no significant difference in HCC development rates between the two groups. Serum AFP levels decreased to similar levels between PEG-IFN/RBV and DCV/ASV groups following the achievement of viral eradication. The risk for HCC development following viral eradication by IFN-free DAA therapy may be similar to that in IFN-based therapy.

ITPA polymorphism effects on decrease of hemoglobin during sofosbuvir and ribavirin combination treatment for chronic hepatitis C

BACKGROUND

Polymorphisms in the inosine triphosphatase (ITPA) gene is associated with anemia induced by peg-interferon (PEG-IFN) plus ribavirin (RBV) treatment for patients with chronic hepatitis C virus (HCV) infection. However, the effect of ITPA polymorphism on sofosbuvir plus RBV treatment is unknown.

METHODS

Two hundred and forty-four patients with chronic HCV genotype 2 infection without decompensated liver cirrhosis were treated with sofosbuvir plus RBV for 12 weeks. The effects of ITPA polymorphism on hemoglobin levels and RBV dose reduction and treatment response were analyzed. ITPA (rs1127354) was genotyped using the Invader assay. Multivariate regression analysis was performed to identify factors associated with sustained virological response (SVR).

RESULTS

Overall, SVR12 was achieved in 231 (94.7%) patients, based on intention to treat analysis. During the therapy, reduction of hemoglobin levels was significantly greater in ITPA genotype CC patients than CA/AA patients. Therefore, the cumulative proportion of patients with RBV dose reduction was significantly higher and total dose of RBV was significantly lower in patients with CC genotype compared to CA/AA genotypes. SVR12 rates were similar between ITPA genotypes CC and CA/AA (94.7 and 94.4%, respectively, P = 0.933). Multivariate logistic regression analysis identified FIB4 index <3.25 (odds ratio [OR], 9.388 for ≥3.25; P = 0.005) and low body weight (OR, 1.059, for high body weight; P = 0.017) as independent predictors for SVR12.

CONCLUSIONS

ITPA polymorphism influences hemoglobin levels and incidence of RBV dose reduction during sofosbuvir plus RBV therapy. However, ITPA genotype CC patients can expect a curative effect equivalent to CA/AA patients for chronic HCV genotype 2 infection.

Thrombopoietin Receptor Agonists Are Often Ineffective in Immune Thrombocytopenia and/or Cause Adverse Reactions: Results from One Hand

BACKGROUND

Eltrombopag and romiplostim are thrombopoietin receptor agonists (TPOs) that have been increasingly used for the treatment of immune thrombocytopenia (ITP). Based on our experience, the incidence of abortive treatment with these drugs and the occurrence of adverse reactions that lead to therapy break-off despite response are higher than has been previously suggested.

METHODS

During the last 8 years, a total of 65 patients were treated with eltrombopag and/or romiplostim at our institute.

RESULTS

36 of a total of 58 patients responded well to eltrombopag. In 12 patients that responded, treatment with eltrombopag was discontinued due to the development of adverse reactions. Eltrombopag was replaced by romiplostim in 23 cases (14 non-responders, 9 patients with adverse reactions). Of these patients, 83% responded to romiplostim. Among all patients treated with romiplostim (n = 32), 75% initially responded; however, 8 of these patients developed adverse reactions. Romiplostim was replaced by eltrombopag in 5 cases (4 due to adverse reactions, 1 non-responsive patient), and only 3 (60%) of these patients were observed to respond to eltrombopag.

CONCLUSION

TPOs often remain ineffective in ITP or result in adverse reactions, which lead to treatment stop or to drug switch. Therefore, alternative treatment options are required.

Outcomes of laparoscopic hepatic resection versus percutaneous radiofrequency ablation for hepatocellular carcinoma located at the liver surface: A case-control study with propensity score matching

AIM

Percutaneous radiofrequency ablation (P-RFA) therapy is a widely applied treatment for small hepatocellular carcinoma (HCC); however, local recurrence is a major issue of HCC located at the surface of the liver (surface HCC). The aim of this study was to compare the outcome of laparoscopic hepatic resection (LH) and P-RFA for surface HCC in case-control patient groups using the propensity score.

METHODS

Between 2011 and 2013, 40 and 52 patients underwent LH and P-RFA for surface HCC (≤3 cm, 1-3 nodules). To correct the difference in clinicopathological factors between the two groups, propensity score matching was used at a 1:1 ratio, which resulted in a comparison of 27 patients/group. We compared outcomes between the two groups, with special reference to local recurrence.

RESULTS

Clinicopathological variables were well balanced between the two groups. One patient in the LH group was converted to open surgery due to adhesion. The incidence of complications was 0% in the P-RFA group and 15% (four patients) in the LH group (P = 0.11); however, none of these four patients in the LH group sustained severe complications. The duration of hospitalization following treatment was longer in the LH group than in the P-RFA group (12.6 vs 7.6 days, P < 0.01). The incidence of local recurrence was lower in the LH group (0%) than in the P-RFA group (eight patients [30%], P = 0.004).

CONCLUSION

LH is an effective treatment for surface HCC with regard to control of local recurrence.

Preoperative diagnosis with versus without MRI in resection for hepatocellular carcinoma

PURPOSE

Although MRI has been considered one of the most sensitive diagnostic techniques for hepatocellular carcinoma (HCC), a clear-cut beneficial effect of the use of preoperative MRI remains unclear. We assessed whether preoperative MRI has a beneficial effect on outcomes in patients scheduled to undergo resection of HCC.

METHODS

We evaluated 449 patients with 553 liver tumors. MRI was performed in 349 of these patients, but not in the other 100. Ultrasonography, dynamic CT, and angiography were performed in all patients. Diagnostic abilities and long-term outcomes were compared between patients who did and did not undergo MRI.

RESULTS

The MRI group (349 patients) had 419 liver tumors and the no MRI group (100 patients) had 134 tumors. Preoperatively, the size of the HCC did not differ between the MRI (median, 30 mm; range, 10-205) and the no MRI group (median, 34 mm; range, 10-175; P = .99). The diagnostic accuracy was 98% in the MRI group and 96% in the no MRI group. Recurrence-free survival rates at 5 years were 31% (95% CI, 20.9-42.5) in the no MRI group, compared with 26% (95% CI, 20.1-32.1) in the MRI (P = .45). Overall survival rates at 5 years were 57% (95% CI, 45.6-68.1) in the no MRI group and 60% (95% CI, 53.4-66.8) in the MRI group (P = .64). After analysis by propensity score matching in 100 pairs of patients, recurrence-free survival rates at 5 years were 31% (95% CI, 20.9-42.5) in the no MRI group, compared with 19% (95% CI, 10.3-30.9) in the MRI group (P = .54). Overall survival rates at 5 years were 57% (95% CI, 45.6-68.1) in the no MRI group and 57% (95% CI, 43.2-68.8) in the MRI group (P = .92).

CONCLUSION

MRI seemed to offer no beneficial impact on diagnostic abilities or long-term outcomes after resection for HCC and is thus of questionable value as a routine imaging modality when combined with CT and angiography clinical practice.

Blood transfusion does not affect survival of gastric cancer patients

BACKGROUND

To initially assess the impact of perioperative blood transfusions (PBTs) on overall survival of patients underwent curative resection of Ⅰ-Ⅲ TNM stage gastric cancer (GC) using the propensity scoring method.

METHODS

The medical records of 1150 GC patients who underwent curative resection in the Tianjin Cancer Hospital between 2003 and 2008 were retrospectively analyzed. Both transfusion and nontransfusion patients were assessed the prognostic differences after surgery using the propensity score analysis.

RESULTS

A total of 299 GC patients (26.0%) were administrated the PBT. With the unadjusted analysis, patients with PBT presented older age, more operative blood loss, lower hemoglobin, lower albumin level, and higher risk of the advanced disease. The 5-y survival rate for patients with PBT was 31.0%, which was significantly lower than that (47.9%) of patients without PBT (P < 0.05). However, we demonstrated that there was not any statistical 5-y survival rate difference of between patients with PBT and patients without PBT with the propensity score analysis (31.0% versus 31.3%, P > 0.05). In addition, we also found that PBT was not significantly associated with the increasing risk of mortality (hazard ratio, 1.054; P = 0.628).

CONCLUSIONS

PBT could not give rise to the worse prognoses of GC patients.

Administration of antithrombin III attenuates posthepatectomy liver failure in hepatocellular carcinoma

BACKGROUND/AIMS

Coagulopathy can cause disseminated intravascular coagulation and posthepatectomy liver failure. Posthepatectomy liver failure predicts a poor prognosis after hepatectomy for hepatocellular carcinoma. Although antithrombin III reduces hypercoagulation, the impact of postoperative antithrombin III administration remains unknown. The aim of this study was to determine whether postoperative antithrombin III administration protects against the development of coagulation disorders.

METHODS

Data from 164 patients who received antithrombin III and 169 who did following curative hepatectomy for hepatocellular carcinoma were retrospectively collected and analyzed. To overcome bias due to different distributions of covariates for the two groups, a one-to-one match was created using propensity score analysis. After matching, patient outcomes were analyzed.

RESULTS

A multivariate analysis of the whole group revealed that antithrombin III activity of <50% on postoperative day 1 was an independent risk factor for posthepatectomy liver failure. After one-to-one matching, the rate of posthepatectomy liver failure was significantly lower in the AT-III-treated group than in the non-AT-III-treated group (16.3% (7/43) vs. 44.2% (19/43), p < 0.01).

CONCLUSIONS

Antithrombin III may attenuate posthepatectomy liver failure in hepatocellular carcinoma, possibly by suppressing coagulopathy.

Clinicopathological features of hepatocellular carcinoma in patients with nonalcoholic fatty liver disease

PURPOSE

The incidence of hepatocellular carcinoma (HCC) in patients with nonalcoholic fatty liver disease (NAFLD) is increasing. However, the clinicopathological features of HCC in these patients are little known. Thus, we investigated the differences in the clinical and pathological characteristics of HCC between NAFLD patients and hepatitis-C virus (HCV) patients.

METHODS

Data from 21 HCC patients with NAFLD and 645 HCC patients with HCV who underwent curative hepatectomy were collected and analyzed. To overcome bias due to differences in the distribution of covariates between the two groups, propensity score matching was performed, and clinicopathological features and outcomes were compared.

RESULTS

In propensity score analysis, the rate of microscopic vascular invasion was significantly higher in the NAFLD group than in the HCV group (65 vs. 30%; P = 0.027). However, overall survival and disease-free survival did not differ between the two matched groups.

CONCLUSIONS

NAFLD may have permissive microenvironment for HCC progression.

One-millimeter cancer-free margin is curative for colorectal liver metastases: a propensity score case-match approach

OBJECTIVE

To investigate the influence of clear surgical resection margin width on disease recurrence rate after intentionally curative resection of colorectal liver metastases.

BACKGROUND

There is consensus that a histological positive resection margin is a predictor of disease recurrence after resection of colorectal liver metastases. The dispute, however, over the width of cancer-free resection margin required is ongoing.

METHODS

Analysis of observational prospectively collected data for 2715 patients who underwent primary resection of colorectal liver metastases from 2 major hepatobiliary units in the United Kingdom. Histological cancer-free resection margin was classified as positive (if cancer cells present at less than 1 mm from the resection margin) or negative (if the distance between the cancer and the margin is 1 mm or more). The negative margin was further classified according to the distance from the tumor in millimeters. Predictors of disease-free survival were analyzed in univariate and multivariate analyses. A case-match analysis by a propensity score method was undertaken to reduce bias.

RESULTS

A 1-mm cancer-free resection margin was sufficient to achieve 33% 5-year overall disease-free survival. Extra margin width did not add disease-free survival advantage (P > 0.05). After the propensity case-match analysis, there is no statistical difference in disease-free survival between patients with negative narrow and wider margin clearance [hazard ratio (HR) 1.0; 95% (confidence interval) CI: 0.9-1.2; P = 0.579 at 5-mm cutoff and HR 1.1; 95% CI: 0.96-1.3; P = 0.149 at 10-mm cutoff]. Patients with extrahepatic disease and positive lymph node primary tumor did not have disease-free survival advantage despite surgical margin clearance (9 months for <1-mm vs 12 months for ≥1-mm margin clearance; P = 0.062).

CONCLUSION

One-mm cancer-free resection margin achieved in patients with colorectal liver metastases should now be considered the standard of care.

5 ways statistics can fool you--tips for practicing clinicians

Published literature suggests that many clinicians are not fully equipped to evaluate and apply research reports for the care of their patients. In this article, we introduce and illustrate five basic statistical concepts that can significantly impact the interpretation of the medical literature and its application to the care of patients, drawing examples from the vaccine literature: (i) consider clinical and statistical significance separately, (ii) evaluate absolute risks rather than relative risks, (iii) examine confidence intervals rather than p values, (iv) use caution when considering isolated significant p values in the setting of multiple testing, and (v) keep in mind that statistically nonsignificant results may not exclude clinically important benefits or harms. These tips may help busy clinicians better interpret the increasingly overwhelming amount of medical literature they are faced with in their daily practices.

Prognostic significance of antithrombin III levels for outcomes in patients with hepatocellular carcinoma after curative hepatectomy

BACKGROUND

Although several studies have shown that serum antithrombin III (ATIII) has anti-inflammatory effects, the prognostic value of ATIII in HCC is unknown. We investigated the influence of preoperative ATIII levels on the outcome of patients who underwent hepatectomy for hepatocellular carcinoma (HCC).

METHODS

Data from 440 patients (314 patients with ATIII ≥ 70 % and 126 patients with ATIII <70 %) who underwent curative hepatectomy for HCC were retrospectively collected and analyzed. To overcome bias due to the different distribution of covariates for the 2 groups, propensity score matching was performed on the patients, and outcomes were compared.

RESULTS

The propensity score analysis revealed that 65 patients with ATIII of ≥ 70 % (group 1) and 65 patients with ATIII of <70 % (group 2) had the same preoperative and operative characteristics (excluding the ATIII level). The overall survival rate and the disease-free survival rate was significantly higher in group 1 than in group 2 (P = 0.005 and 0.011, respectively). Multivariate analysis showed that ATIII was a significant favorable factor for overall survival and disease-free survival of patients with HCC after curative hepatectomy.

CONCLUSIONS

The prognosis of patients with HCC was found to be associated with preoperative antithrombin III levels. ATIII may be useful for predicting outcomes of patients with HCC after curative hepatectomy.

Scientific rigour in psycho-oncology trials: why and how to avoid common statistical errors

OBJECTIVE

It is well documented that statistical and methodological flaws are common in much of the health research literature, including psycho-oncology. These can have far-reaching effects, including the publishing of misleading results; the wasting of time, effort, and financial resources; exposure of patients to the potential harms of research and decreased confidence in science and researchers by the public.

METHODS

Several of the most common statistical errors and methodological pitfalls that occur in the field of psycho-oncology are discussed, including those that occur at the design, analysis, reporting and conclusion stages.

RESULTS

Fourteen topics are briefly discussed, explaining why there is a problem and how to avoid it. These include proper approaches to power, clustering, missing data, categorization of continuous variables, subgroup analyses, multiple comparisons, statistical interactions, confidence intervals and correct interpretation of p-values. Extensive referencing points the reader to more in-depth explanations.

CONCLUSIONS

To increase the scientific rigour in psycho-oncology, researchers should involve a biostatistician from the beginning of the study and should commit to continuing education on best practices in the fields of statistics and reporting.

Safety of hepatic resection for colorectal metastases in the era of neo-adjuvant chemotherapy

PURPOSE

The relationship between neo-adjuvant chemotherapy prior to hepatectomy in patients with resectable colorectal liver metastases and post-operative morbidity still has to be clarified.

METHODS

Data from 242 patients undergoing hepatectomy for colorectal liver metastases, judged resectable at first observation, were reviewed and their clinical outcome was related to neo-adjuvant chemotherapy (125 patients). Selection biases were outlined and properly handled by means of propensity score analysis.

RESULTS

Post-operative death was 1.2% and morbidity 40.9%. Pre-operative chemotherapy was only apparently related to higher morbidity (P = 0.021): multivariate analysis identified extension of hepatectomy and intra-operative blood loss as independent prognostic variables (P < 0.05). Patients receiving and not receiving neo-adjuvant chemotherapy were significantly different for several covariates, including extension of hepatectomy (P = 0.049). After propensity score adjustment, 94 patients were identified as having similar covariate distribution (standardized differences <|0.1|) except for neo-adjuvant treatment (47 patients for each group). In this matched sample, mortality was similar and post-operative complications were only slightly higher (hazard ratio = 1.38) in treated patients. A significantly higher need for fluid replacement was only observed in patients receiving neo-adjuvant chemotherapy (P = 0.038).

CONCLUSIONS

Neo-adjuvant chemotherapy showed a limited role in determining post-operative morbidity after hepatic resection and did not modify mortality.

Impact of pegylated interferon therapy on outcomes of patients with hepatitis C virus-related hepatocellular carcinoma after curative hepatic resection

BACKGROUND

Several published reports investigating the effects of interferon (IFN) therapy on survival and tumor recurrence after curative resection of hepatocellular carcinoma (HCC) have been inconclusive. The aim of this study is to investigate the efficacy of pegylated-IFN (peg-IFN) therapy after curative hepatic resection for HCC in patients infected with hepatitis C virus (HCV).

METHODS

Data from 175 patients who underwent curative hepatic resection for HCC associated with HCV were retrospectively collected and analyzed; 75 patients received peg-IFN therapy after surgery, whereas 100 patients did not receive IFN therapy. To overcome biases resulting from the different distribution of covariates in the two groups, a one-to-one match was created using propensity score analysis. After matching, patient outcomes were analyzed.

RESULTS

After one-to-one matching, patients (n = 38) who received peg-IFN therapy after surgery and patients (n = 38) who did not receive IFN therapy had the same preoperative and operative characteristics. The 3- and 5-year overall survival rates of patients who received peg-IFN therapy after hepatic resection were significantly higher than those of patients who did not receive IFN therapy (P = 0.00135). The 3- and 5-year overall survival rates were 100 and 91.7% and 76.6 and 50.6% in the peg-IFN group and non-IFN group, respectively. There was no significant difference in disease-free survival between the two matched groups (P = 0.886).

CONCLUSION

Peg-IFN therapy may be effective as an adjuvant chemopreventive agent after hepatic resection in patients with HCV-related HCC.

Systematic survey of the design, statistical analysis, and reporting of studies published in the 2008 volume of the Journal of Cerebral Blood Flow and Metabolism

Translating experimental findings into clinically effective therapies is one of the major bottlenecks of modern medicine. As this has been particularly true for cerebrovascular research, attention has turned to the quality and validity of experimental cerebrovascular studies. We set out to assess the study design, statistical analyses, and reporting of cerebrovascular research. We assessed all original articles published in the Journal of Cerebral Blood Flow and Metabolism during the year 2008 against a checklist designed to capture the key attributes relating to study design, statistical analyses, and reporting. A total of 156 original publications were included (animal, in vitro, human). Few studies reported a primary research hypothesis, statement of purpose, or measures to safeguard internal validity (such as randomization, blinding, exclusion or inclusion criteria). Many studies lacked sufficient information regarding methods and results to form a reasonable judgment about their validity. In nearly 20% of studies, statistical tests were either not appropriate or information to allow assessment of appropriateness was lacking. This study identifies a number of factors that should be addressed if the quality of research in basic and translational biomedicine is to be improved. We support the widespread implementation of the ARRIVE (Animal Research Reporting In Vivo Experiments) statement for the reporting of experimental studies in biomedicine, for improving training in proper study design and analysis, and that reviewers and editors adopt a more constructively critical approach in the assessment of manuscripts for publication.

Biostatistics: a toolkit for exploration, validation, and interpretation of clinical data

Biostatistics plays a key role in all phases of clinical research starting from the design to the monitoring, data collection, data analysis, and interpretation of the results. A clear understanding of the statistical framework as it relates to the study hypothesis, reported results, and interpretation is vital for the scientific integrity of the study and its acceptance in the general medical community. In this brief report, we will put in perspective the general analytical framework for exploring and validating prognostic factors using data from large databases.

Is portal hypertension a contraindication to hepatic resection?

BACKGROUND AND AIMS

The outcome of hepatic resection in cirrhotic patients has improved remarkably in recent years with improved surgical techniques and perioperative care; however, the role of portal hypertension is still uncertain. The aim of this study was to elucidate surgical outcomes of hepatectomy in patients with portal hypertension.

METHODS

Data from 241 cirrhotic patients who underwent resection for hepatocellular carcinoma were retrospectively collected and analyzed: patients were divided into 2 groups according to the presence (n = 89) or absence (n = 152) of portal hypertension at the time of surgery. To overcome biases owing to the different distribution of covariates throughout the 2 groups, a one-to-one match was created using propensity score analysis: after match, intraoperative, and postoperative course and survival rates were analyzed.

RESULTS

Patients with portal hypertension experienced worse preoperative liver function (mean model for end-stage liver disease [MELD] score, 9.5 +/- 7.8 vs. 8.4 +/- 1.3; P = 0.001) and survival rates (P = 0.008) in comparison to those without portal hypertension: after one-to-one matching, patients with (n = 78) and without portal hypertension (n = 78) had the same preoperative characteristics and showed the same intraoperative course, postoperative occurrence of liver failure, morbidity, length of in-hospital stay and survival rates (P = ns in all cases). The only predictors of postoperative liver failure were MELD score (P = 0.001) and extent of hepatectomy (P = 0.005).

CONCLUSIONS

Faced with the same MELD score and extent of hepatectomy planning, presence of portal hypertension should not be considered as a contraindication for hepatic resection in cirrhotic patients.

Proposed scoring system to determine small bowel mass lesions using capsule endoscopy

BACKGROUND/PURPOSE

Capsule endoscopy is a highly sensitive method for the detection of small bowel lesions. False-positive findings are important confounding factors. This study reports on a scoring system for evaluating the role of capsule endoscopy in small bowel tumors.

METHODS

Ten men and two women (age, 23-79 years) with suspected small bowel tumors were included from 120 patients referred for capsule endoscopy between March 2004 and March 2008. The indications were gastrointestinal bleeding (n = 9), melanoma workup (n = 1), physical checkup (n = 1), and iron deficiency anemia (n = 1). The proposed tumor score was composed of five components: bleeding, mucosal disruption, an irregular surface, color, and white villi. These can be scored for probability of mass lesions seen at capsule endoscopy. Small bowel mass lesions were probably present in those with a score of >or= 4, and a score of <or= 2 indicated a low probability of a small bowel mass lesion.

RESULTS

Capsule endoscopy showed probable small bowel mass lesions in six patients, and a low lesion probability in the other six. Capsule endoscopy showed that new lesions were not detected by esophagogastroduodenoscopy or colonoscopy. All six patients with probable small bowel tumors were found to have pathological findings upon capsule endoscopy: two with lymphangioma, and one each with ileal ectopic pancreas, with melanoma metastasis, gastrointestinal lymphoma, and gastrointestinal stromal tumor.

CONCLUSION

Capsule endoscopy may detect small bowel tumors more reliably by using the scoring system outlined. It should be considered in suspected cases of small bowel tumor.