Agreement between measurements refers to the degree of concordance between two (or more) sets of measurements. Statistical methods to test agreement are used to assess inter-rater variability or to decide whether one technique for measuring a variable can substitute another. In this article, we look at statistical measures of agreement for different types of data and discuss the differences between these and those for assessing correlation.

In a previous article in this series, we explained correlation analysis which describes the strength of relationship between two continuous variables. In this article, we deal with linear regression analysis which predicts the value of one continuous variable from another. We also discuss the assumptions and pitfalls associated with this analysis.

During the conduct of clinical trials, it is not uncommon to have protocol violations or inability to assess outcomes. This article in our series on common pitfalls in statistical analysis explains the complexities of analyzing results from such trials and highlights the importance of “intention-to-treat” analysis.

In clinical research, study results, which are statistically significant are often interpreted as being clinically important. While statistical significance indicates the reliability of the study results, clinical significance reflects its impact on clinical practice. The third article in this series exploring pitfalls in statistical analysis clarifies the importance of differentiating between statistical significance and clinical significance.

Background The COVID-19 pandemic has disrupted health-care systems, leading to concerns about its subsequent impact on non-COVID disease conditions. The diagnosis and management of cancer is time sensitive and is likely to be substantially affected by these disruptions. We aimed to assess the impact of the COVID-19 pandemic on cancer care in India. MethodsWe did an ambidirectional cohort study at 41 cancer centres across India that were members of the National Cancer Grid of India to compare provision of oncology services between March 1 and May 31, 2020, with the same time period in 2019. We collected data on new patient registrations, number of patients visiting outpatient clinics, hospital admissions, day care admissions for chemotherapy, minor and major surgeries, patients accessing radiotherapy, diagnostic tests done (pathology reports, CT scans, MRI scans), and palliative care referrals. We also obtained estimates from participating centres on cancer screening, research, and educational activities (teaching of postgraduate students and trainees). We calculated proportional reductions in the provision of oncology services in 2020, compared with 2019. FindingsBetween March 1 and May 31, 2020, the number of new patients registered decreased from 112 270 to 51 760 (54% reduction), patients who had follow-up visits decreased from 634 745 to 340 984 (46% reduction), hospital admissions decreased from 88 801 to 56 885 (36% reduction), outpatient chemotherapy decreased from 173634 to 109 107 (37% reduction), the number of major surgeries decreased from 17 120 to 8677 (49% reduction), minor surgeries from 18 004 to 8630 (52% reduction), patients accessing radiotherapy from 51 142 to 39 365 (23% reduction), pathological diagnostic tests from 398 373 to 246 616 (38% reduction), number of radiological diagnostic tests from 93 449 to 53 560 (43% reduction), and palliative care referrals from 19 474 to 13 890 (29% reduction). These reductions were even more marked between April and May, 2020. Cancer screening was stopped completely or was functioning at less than 25% of usual capacity at more than 70% of centres during these months. Reductions in the provision of oncology services were higher for centres in tier 1 cities (larger cities) than tier 2 and 3 cities (smaller cities).Interpretation The COVID-19 pandemic has had considerable impact on the delivery of oncology services in India. The long-term impact of cessation of cancer screening and delayed hospital visits on cancer stage migration and outcomes are likely to be substantial.

Background Whether current postoperative surveillance regimes result in improved overall survival (OS) of patients with extremity sarcomas is unknown. Questions/purposes We hypothesized that a less intensive followup protocol would not be inferior to the conventional followup protocol in terms of OS. We (1) assessed OS of patients to determine if less intensive followup regimens led to worsened survival and asked (2) whether chest radiograph followup group was inferior to CT scan followup group in detecting pulmonary metastasis; and (3) whether less frequent (6-monthly) followup interval was inferior to more frequent (3-monthly) followup in detecting pulmonary metastasis and local recurrence.Methods A prospective randomized single-center noninferiority trial was conducted between January 2006 and June 2010. On the basis of 3-year survival of 60% with intensive, more frequent followup, 500 nonmetastatic patients were randomized to demonstrate noninferiority by a margin (delta) of 10% (hazard ratio [HR], 1.36). The primary end point was OS at 3 years. The secondary objective was to compare disease-free survival (DFS) (time to recurrence) at 3 years. At minimum followup of 30 months (median, 42 months; range, 30-81 months), 178 deaths were documented. Results Three-year OS and DFS for all patients was 67% and 52%, respectively. Three-year OS was 67% and 66% in chest radiography and CT groups, respectively (HR, 0.9; upper 90% confidence interval [CI], 1.13). DFS rate was 54% and 49% in chest radiography and CT groups, respectively (HR, 0.82; upper 90% CI, 0.97). Three-year OS was 64% and 69% in 6-monthly and 3-monthly groups, respectively (HR, 1.2; upper 90% CI,

In biomedical research, we are often interested in quantifying the relationship between an exposure and an outcome. “Odds” and “Risk” are the most common terms which are used as measures of association between variables. In this article, which is the fourth in the series of common pitfalls in statistical analysis, we explain the meaning of risk and odds and the difference between the two.

No abstract