Repeatability

It is a measure of precision. It is the «variation due to the measuring device. It is the variation observed when the same operator measures the same part repeatedly with the same device [1].» It «is similar to uniformity except that it deals with how consistent a particular sensor is against itself. It can be used to describe the ability of a sensor to provide the same result, under the same circumstances, over and over again [2].»

Bibliographic references:

[1] Reliawiki. 2017. Measurement System Analysis. [online] Available at: <https://reliawiki.org/index.php/Measurement_System_Analysis> [Accessed 20 February 2022].

[2] Apogeeinstruments.com. 2022. Uniformity, Repeatability, Stability, and Accuracy. [online] Available at: <https://www.apogeeinstruments.com/uniformity-repeatability-stability-and-accuracy/> [Accessed 20 February 2022].

Stability

It is «a measure of how the accuracy and precision of the system perform over time [1].»

[1]

«Stability deals with the degree to which sensor characteristics remain constant over time. Changes in stability, also known as drift, can be due to components aging, decrease in sensitivity of components, and/or a change in the signal to noise ratio, etc [2].»

Bibliographic references:

[1] Reliawiki. 2017. Measurement System Analysis. [online] Available at: <https://reliawiki.org/index.php/Measurement_System_Analysis> [Accessed 20 February 2022].

[2] Apogeeinstruments.com. 2022. Uniformity, Repeatability, Stability, and Accuracy. [online] Available at: <https://www.apogeeinstruments.com/uniformity-repeatability-stability-and-accuracy/> [Accessed 20 February 2022].

Precision

«Precision describes the variation you see when you measure the same part repeatedly with the same device. It includes the following two types of variation:

• Repeatability
• Reproducibility: variation due to the operators and the interaction between operator and part. It is the variation of the bias observed when different operators measure the same parts using the same device.»
  

Bibliographic reference: Reliawiki. 2017. Measurement System Analysis. [online] Available at: <https://reliawiki.org/index.php/Measurement_System_Analysis> [Accessed 20 February 2022].

Accuracy

«Accuracy describes the difference between the measurement and the actual value of the part that is measured. It includes:

• Bias: a measure of the difference between the true value and the observed value of a part. If the “true” value is unknown, it can be calculated by averaging several measurements with the most accurate measuring equipment available.
• Linearity: a measure of how the size of the part affects the bias of a measurement system. It is the difference in the observed bias values through the expected range of measurement [1].»

  

  [1]

It is «analogous to uncertainty relative to a reference, is in its simplest terms the difference between measured and “true” values [2].»

Bibliographic references:

[1] Reliawiki. 2017. Measurement System Analysis. [online] Available at: <https://reliawiki.org/index.php/Measurement_System_Analysis> [Accessed 20 February 2022].

[2] Apogeeinstruments.com. 2022. Uniformity, Repeatability, Stability, and Accuracy. [online] Available at: <https://www.apogeeinstruments.com/uniformity-repeatability-stability-and-accuracy/> [Accessed 20 February 2022].

Actuarial (versus actual)

«Its use is often extended (...) to include the Kaplan-Meier (KM) method, (...). Actual is a new term (...) to refer to a technique with several designations in the statistical literature, including cumulative incidence and crude, unadjusted, absolute, or observable probability [1].»

«Actuarial analysis, (...), is used to describe and compare survival probabilities by allowing for partial survival times (censoring). (...) The actuarial event-free (survival) curve for a nonfatal event, such as structural valve deterioration (SVD) of porcine valves, estimates the event-free probability for a population in which death has been eliminated. This overestimates the percentage of valves that will actually fail, because many patients die before the valve fails [2].»

«Unlike cumulative incidence [actual estimates], the KM attempts to predict what the latent failure probability would be if death were eliminated. To do this, the KM method assumes that the risk of dying and the risk of failure are independent. But this assumption is not true for many cardiac applications in which the risks of failure and death are negatively correlated (ie, patients with a higher risk of dying have a lower risk of failure, and patients with a lower risk of death have a higher risk of failure, which is a condition called informative censoring) [3].»

«When used for nonfatal events such as SVD, actual analysis estimates the percentage of patients who will have SVD (or the probability that an individual patient will experience SVD). The KM (actuarial) method attempts to estimate the percentage of SVD that would occur if patients never died. Also, KM depends on the assumption that death and SVD are independent, which they are probably not [1].»

Bibliographic references:

[1] Grunkemeier GL, Wu Y. Actual versus actuarial event-free percentages. Ann Thorac Surg. 2001 Sep;72(3):677-8. Available at: https://doi.org/10.1016/s0003-4975(01)03059-4.

[2] Grunkemeier GL, Jamieson WR, Miller DC, Starr A. Actuarial versus actual risk of porcine structural valve deterioration. J Thorac Cardiovasc Surg. 1994 Oct;108(4):709-18. PMID: 7934107. Available at: https://doi.org/10.1016/S0022-5223(94)70298-5.

[3] Grunkemeier GL, Jin R, Eijkemans MJ, Takkenberg JJ. Actual and actuarial probabilities of competing risks: apples and lemons. Ann Thorac Surg. 2007 May;83(5):1586-92. Available at: https://doi.org/10.1016/j.athoracsur.2006.11.044.

Employing medical writers & frequentist vs bayesian methods with Dr. Allen Pannell

Liston to: https://soundcloud.com/plenarysession/166-employing-medical-writers-frequentist-vs-bayesian-methods-with-dr-allen-pannell

Image source: Rosiehilder. 12 Terrific Ways to Practice Your Spanish Listening Skills. FluentU Spanish. https://www.fluentu.com/blog/spanish/spanish-listening-comprehension-practice-online/. Accessed December 28, 2019.

Cancer-related survival

It is the time from diagnosis of cancer, or the start of treatment for cancer, to the date of death related to primary cancer.

Bibliographic reference: Wu YC, et al. Long-term results of pathological stage I non-small cell lung cancer: validation of using the number of totally removed lymph nodes as a staging control. Eur J Cardiothorac Surg. 2003 Dec;24(6):994-1001. Available at: https://doi.org/10.1016/S1010-7940(03)00567-0.

P value

Source: Twitter.com. (2019). Twitter. [online] Available at: https://twitter.com/drnirmalraut/status/1090550945593856000?s=09 [Accessed 3 Feb. 2019].

Follow-up

«Monitoring a person's health over time after treatment. This includes keeping track of the health of people who participate in a clinical study or clinical trial for a period of time, both during the study and after the study ends [1]», «(...) in order to observe changes in health status or health-related variables [2].»

Bibliographic references:

[1] National Cancer Institute. NCI Dictionary of Cancer Terms. https://www.cancer.gov/publications/dictionaries/cancer-terms/def/follow-up. Accessed November 21, 2018.

[2] Last J. A Dictionary Of Epidemiology. 4th ed. Oxford: Oxford University Press; 2001:72.

Disease-free survival (DFS) or relapse-free survival (RFS)

In cancer, the length of time after the end of primary treatment for cancer and that the patient survives without any signs or symptoms of that cancer. In a clinical trial, measuring the DFS is one way to see how well a new treatment works [1].

It includes local, regional, or distant recurrence and death due to any cause.

Bibliographic references:

[1] National Cancer Institute. NCI Dictionary of Cancer Terms. Available at: https://www.cancer.gov/publications/dictionaries/cancer-terms/def/dfs. Accessed November 21, 2018.

Overall survival (OS)

«The length of time from either the date of diagnosis or the start of treatment for a disease, such as cancer, that patients diagnosed with the disease are still alive [1].» It is «the process of locating research subjects or patients to determine whether or not some outcome of interest has occurred [2].»  It includes any cause of death.

Bibliographic references:
[1] National Cancer Institute. NCI Dictionary of Cancer Terms. Available at: https://www.cancer.gov/publications/dictionaries/cancer-terms/def/overall-survival. Accessed November 21, 2018.
[2] Everitt B, Skrondal A. The Cambridge Dictionary Of Statistics. 4th ed. Cambridge: Cambridge University Press; 2011:170.

It’s all about statistics:

Source: Carvalho R. Twitter. Twitter.com. https://twitter.com/rcarvalhoonco/status/1045489255445934082?s=03. Published 2018. Accessed October 17, 2018.

Statistics

«The statistics studies how to collect data (How many? In what way?) and how to analyze them to obtain the information that allows answering the questions that we put. It is about producing knowledge through observation and analysis of reality, in an intelligent and objective way. It is the essence of the scientific method.»
«To summarize a set of information from a large dataset, certain quantities are selected which concentrate the most possible information, and through them - can be 5 or 6 points - we can get a fairly accurate idea of the behavior of the data in general. These amounts are normally divided into three groups: tendency (center), scale (or dispersion), and location.»

Bibliographic reference: Grima P. Os segredos da estatística - a certeza absoluta e outras ficções. National Geographic - Edição Especial Matemática. 2018:15.

Basket study

Source: [4].

«Basket trials (or studies) test the effect of one drug on a single mutation in a variety of tumor types [baskets], at the same time. These studies also have the potential to greatly increase the number of patients who are eligible to receive certain drugs relative to other trials designs [1], «allowing researchers to analyze each cancer type individually, as well as assess the impact of the drug or drug combinations as a whole. Using this approach, it is possible to combine what would have been multiple phase 2 trials into a single study [4]». «In the evaluation of targeted therapies, basket trials have emerged as an approach to test the hypothesis that targeted therapies may be effective independent of tumor histology, as long as the molecular target is present [2].» «Most basket trials typically aim to answer multiple questions simultaneously [3].»

Bibliographic references:
[1] Clinical Trial Design and Methodology. ASCO. https://www.asco.org/research-progress/clinical-trials/clinical-trial-resources/clinical-trial-design-and-methodology. Accessed August 18, 2018.

[2] Redig AJ, Jänne PA. Basket trials and the evolution of clinical trial design in an era of genomic medicine. J Clin Oncol. 2015 Mar 20;33(9):975-7. Available at: https://doi.org/10.1200/JCO.2014.59.8433.

[3] Cunanan KM, Gonen M, Shen R, et al. Basket Trials in Oncology: A Trade-Off Between Complexity and Efficiency. J Clin Oncol. 2017 Jan 20;35(3):271-273. Available at: https://doi.org/10.1200/JCO.2016.69.9751.

[4] Illuminating ideas: Innovative clinical trial design. Roche.com. https://www.roche.com/research_and_development/who_we_are_how_we_work/clinical_trials/innovative-clinical-trial-design.htm. Accessed August 18, 2018.

Multi-arm multi-stage (MAMS) trial design

Ratio

It's the relationship between two sets with different characteristics. The numerator is not included in the denominator. It is different from proportion.

Proportion

It's the relationship between the number of individuals who have a characteristic and the total population. The numerator is included in the denominator. It is different from ratio.

Missing data or missing values

In statistics, they «occur when no data value is stored for the variable in an observation [1].» They are «values of variables within data sets which are not known [2].»

Bibliographic references:

[1] Missing data [Internet]. En.wikipedia.org. 2016 [cited 19 November 2016]. Available from: https://en.wikipedia.org/wiki/Missing_data.

[2] Statistics Glossary: M [Internet]. Statsoft.com. 2016 [cited 19 November 2016]. Available from: http://www.statsoft.com/textbook/statistics-glossary/m#Missing values.

Standard error (SE)

«These are the SEs for the descriptive statistics. The SE gives some idea about the variability possible in the statistic [1].» It «(...) is a measure of the variability of a statistic. It is an estimate of the standard deviation of a sampling distribution [2].» It «(...) is the standard deviation of the sampling distribution of a statistic, (...) (Everitt BS, 2003, cited in [3]). «The SE of the mean [SEM] (...) is the theoretical standard deviation of all sample means of size n drawn from a population and depends on both the population variance (sigma) and the sample size (n) (...) [4].» «The SEM can be seen to depict the relationship between the dispersion of individual observations around the population mean (the standard deviation), and the dispersion of sample means around the population mean (the SE). Different samples drawn from that same population would in general have different values of the sample mean, so there is a distribution of sampled means (with its own mean and variance). (...) As the sample size increases, the dispersion of the sample means clusters more closely around the population mean and the SE decreases [3].» «The SE depends on three factors: the number of observations in the population (N), the number of observations in the sample (n), and the way that the random sample is chosen. If the population size is much larger than the sample size, then the sampling distribution has roughly the same SE, whether we sample with or without replacement . On the other hand, if the sample represents a significant fraction (say, 1/20) of the population size, the SE will be noticeably smaller, when we sample without replacement [2].» «The SE of the proportion (...) is the standard deviation of the distribution of the sample proportion over repeated samples [4].»

Source: Steve's Favorites. Pinterest. Available at: https://www.pinterest.pt/pin/419819996504206121/. Accessed November 3, 2017.

Bibliographic references:

[1] Annotated SPSS Output: Descriptive statistics [Internet]. Ats.ucla.edu. 2016 [cited 20 November 2016]. Available from: http://www.ats.ucla.edu/stat/spss/output/descriptives.htm.

[2] Statistics Dictionary [Internet]. Stattrek.com. 2016 [cited 20 November 2016]. Available from: http://stattrek.com/statistics/dictionary.aspx.

[3] Standard error [Internet]. En.wikipedia.org. 2016 [cited 20 November 2016]. Available from: https://en.wikipedia.org/wiki/Standard_error#cite_note-1.

[4] Statistics Glossary: S [Internet]. Statsoft.com. 2016 [cited 20 November 2016]. Available from: http://www.statsoft.com/textbook/statistics-glossary/s#Standard.

Confidence interval (CI)

It is «a range of values, calculated from the sample observations, that is believed, with a particular probability, to contain the true parameter value» [1]. It is used «to express the degree of uncertainty associated with a sample statistic. A confidence interval is an interval estimate combined with a probability statement.» One «might describe the interval estimate as a "95% confidence interval". This means that if» one «used the same sampling method to select different samples and computed an interval estimate for each sample,» one «would expect the true population parameter to fall within the interval estimates 95% of the time» [2]. «Precision is taken to be the narrowness of the confidence interval. (...) The interval estimate is an expression of the uncertainty surrounding the point estimate and derives mainly from sampling variation as well as measurement variation/error. In general, the degree of uncertainty is inversely related to the size of the study. On one hand, if a study is too small, the uncertainty may increase to a level considered to be undesirable or useless. On the other, as the study size increases, the degree of uncertainty decreases, and the interval estimate becomes narrower» [3]. «Confidence intervals are preferred to point estimates and to interval estimates, because only confidence intervals indicate the precision of the estimate and the uncertainty of the estimate» [2].

Bibliographic references:

[1] Everitt, B. and Skrondal, A. (2010). The Cambridge dictionary of statistics. 4th ed. Cambridge, UK: Cambridge University Press.

[2] Stattrek.com. (2016). Statistics Dictionary. [online] Available at: http://stattrek.com/statistics/dictionary.aspx [Accessed 13 Nov. 2016].

[3] Broeck, J. and Brestoff, J. (2013). Epidemiology: Principles and Practical Guidelines. 1st ed. Dordrecht: Springer.