The Power of Asterisks

Horowitz, Gary L.

doi:10.1373/clinchem.2015.243048

Cited by 18 publications

(9 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The Canadian Health Measures Survey reports changes in reference intervals mainly in children. While 4 070 samples were available from individuals aged 15–79 years, 79 % of screened individuals aged 60–79 years old were excluded (Adeli et al ., ; Horowitz, ). This makes the judgement of specific changes in the elderly difficult, although, a slight increase of MCV and a decrease of haemoglobin in men can be observed with increasing age (Adeli et al ., ).…”

Section: Discussionmentioning

confidence: 99%

Blood counts in adult and elderly individuals: defining the norms over eight decades of life

Zierk

Krebs²,

Rauh

et al. 2020

Br J Haematol

View full text Add to dashboard Cite

The blood count is one of the most common tests used for health assessment. In elderly individuals, selection of a 'healthy' reference population for laboratory assessment is difficult due to the high prevalence of chronic morbidities, leading to uncertainty regarding appropriate reference intervals. In particular, age-specific lower haemoglobin reference limits to define anaemia are controversial. Here, we applied a data mining approach to a large dataset of 3 029 904 clinical routine samples to establish blood count reference intervals. We excluded samples from units/specialists with a high proportion of abnormal blood counts, samples from patients with an unknown or decreased estimated glomerular filtration rate, and samples with abnormal test results in selected other analytes. After sample exclusion, 566 775-572 060 samples from different individuals aged 20-100 years were available for analysis. We then used an established statistical algorithm to determine the distribution of physiological test results and calculated age-and sex-specific reference intervals. Our results show substantial trends with age in haematology analytes' reference intervals. Most notably, haemoglobin and red cell counts decline in men with advanced age, accompanied by increases in red cell volume in both sexes. These findings were confirmed in an independent dataset, and suggest an at least partly physiologic cause.The blood count is one of the most common tests for health assessment with diagnostic and therapeutic implications for a multitude of common to rare and minor to life-threatening conditions. However, despite its frequent clinical use and importance and near-universal availability on the one hand, and an ageing population throughout the world on the other hand, uncertainty regarding appropriate reference intervals for older individuals still exists. In particular, age-specific lower reference limits for haemoglobin, which define anaemia, are controversial, despite substantial consequences for individual patients and considerable public health impact.A major challenge when establishing reference intervals for individuals >50 years is the increasing proportion of chronic morbidities and medication with age, which leads to exclusion of these subjects from conventional reference interval studies (Adeli et al., 2015;R€ ohrig et al., 2018). Based on these restrictions, data mining of age-specific reference intervals using laboratory test results collected from routine patient care can be considered a viable complement to conventional reference interval studies (Haeckel et al., 2017;Jones et al., 2018). In contrast to the highly selected population of conventional approaches, a reference interval data mining approach uses exactly the 'real world' population to which the reference intervals are ultimately applied, while still excluding outliers. In addition, the necessity to obtain a patient history and perform a clinical examination restricts the number of individuals who can be recruited for a conventional reference sample, while...

show abstract

Section: Discussionmentioning

confidence: 99%

Blood counts in adult and elderly individuals: defining the norms over eight decades of life

Zierk

Krebs²,

Rauh

et al. 2020

Br J Haematol

View full text Add to dashboard Cite

show abstract

“…Laboratory test results need to be accompanied by appropriate reference intervals to support clinical decision-making 1,2 . Conventional protocols for the establishment of reference intervals ("direct methods") require the sampling of a carefully selected, sufficiently large (usually >120 individuals), and homogenous group of healthy reference individuals, and the 2.5 th and 97.5 th percentiles of test results define the reference interval [3][4][5] .…”

mentioning

confidence: 99%

“…Additionally, the use of direct methods results in substantial differences between the reference population and the target population in which the reference intervals are eventually used. Most patients are significantly older, use prescription drugs, and have relevant co-morbidities 1 . This results in uncertainty regarding the suitability of reference intervals based on blood samples from young and healthy adults, and limits the creation of accurate reference intervals for children and elderly adults.…”

mentioning

confidence: 99%

Reference Interval Estimation from Mixed Distributions using Truncation Points and the Kolmogorov-Smirnov Distance (kosmic)

Zierk

Arzideh

Kapsner

et al. 2020

Sci Rep

View full text Add to dashboard Cite

Appropriate reference intervals are essential when using laboratory test results to guide medical decisions. Conventional approaches for the establishment of reference intervals rely on large samples from healthy and homogenous reference populations. However, this approach is associated with substantial financial and logistic challenges, subject to ethical restrictions in children, and limited in older individuals due to the high prevalence of chronic morbidities and medication. We implemented an indirect method for reference interval estimation, which uses mixed physiological and abnormal test results from clinical information systems, to overcome these restrictions. The algorithm minimizes the difference between an estimated parametrical distribution and a truncated part of the observed distribution, specifically, the Kolmogorov-Smirnov-distance between a hypothetical Gaussian distribution and the observed distribution of test results after Box-Cox-transformation. Simulations of common laboratory tests with increasing proportions of abnormal test results show reliable reference interval estimations even in challenging simulation scenarios, when <20% test results are abnormal. Additionally, reference intervals generated using samples from a university hospital's laboratory information system, with a gradually increasing proportion of abnormal test results remained stable, even if samples from units with a substantial prevalence of pathologies were included. A highperformance open-source c++ implementation is available at https://gitlab.miracum.org/kosmic.

show abstract

“…First, research laboratory are a foremost part of today's healthcare organizations. However, despite high throughput with low turnaround times, the capacity to screen data for results of special interest has decreased and few tests are directly diagnostic [20]. Second, technical improvements have enabled the integration of ES capabilities and software presentations, including automatic analyzers and modules of research laboratory information systems [21].…”

Section: Review Of Machine Learning In Medicinementioning

confidence: 99%

“…Third, because patients can now directly access their research laboratory test results of their investigative benefactor, there is an increasing demand for meaningful, possibly personalized reference limits and the need to interpret precision asterisks [20], the conventional signs indicating abnormal or borderline values. Finally, with the convergence of smartphones and innovative biosensors based on microfluidics and microelectronics, the vision of the lab-on-a-chip (LOC) and related models for laboratory medicine has opened opportunities, "in which a smartphoneenabled portable laboratory is brought to the patient instead of the patient being brought to the laboratory" [23].…”

Section: Review Of Machine Learning In Medicinementioning

confidence: 99%