Validated Protocols for the Analysis of Dried Blood Spot Samples
Protocols for the analysis of over 100 analytes in dried blood spot samples have been published, including important indicators of endocrine, immune, reproductive, and metabolic function, as well as measures of nutritional status and infectious disease (McDade et al., 2007). Many of these biomarkers have been applied clinically, and may be used in survey research to determine risk for the development of disease, to gain insight into the impact of psychosocial/behavioral contexts across multiple physiological systems, and/or to understand the effect of health on selection into environments across the life course.
Table 2 (PDF) provides more detailed information for a subset of analytes most likely to be of interest to researchers conducting population-level, community-based health research. This list was originally published in "What a Drop Can Do: Dried Blood Spots as a Minimally Invasive Method for Integrating Biomarkers into Population-Based Research," (PDF) by Thomas McDade, Sharon Williams, and J. Josh Snodgrass, (2007), and inclusion is based on four criteria:
- Methods that use capillary whole blood collected on filter paper, without requiring the separation of erythrocytes or additional sample processing steps.
- Markers of physiological function and health that are broadly relevant across a wide range of ages. The table does not include markers of inborn errors of metabolism commonly used for neonatal screening, markers of toxicology, or clinical markers of specific diseases, unless they are likely to be relevant at the population level (e.g., HIV, hepatitis).
- Evidence of attention to assay performance, including a report of accuracy, precision, reliability, and/or analysis of matched blood spot and serum/plasma samples.
- Publication in a peer-reviewed journal.
Table 2 includes information on multiple aspects of assay performance and implementation to guide decisions regarding the utility and feasibility of various blood spot methods. Table 2 is updated annually. Please email Thom McDade with suggestions for analytes and/or references to add to this list.
Definitions of each heading:
Volume of sample
A typical drop of capillary blood collected from a finger stick includes approximately 50uL of whole blood. Most assays use a hole punch to punch out a disc of dried blood of a given size for analysis, while others use the entire spot. Linear dimensions (i.e., mm or inches) pertain to methods using a hole punch, while volume measures (i.e., uL) are presented for methods that use an entire blood spot containing a pre-measured quantity of whole blood.
There are no standardized criteria for acceptable levels of sample degradation, so we rely on the stability determination as published. In many cases, the reported stability reflects the maximum period of time evaluated, and therefore actual stability may be significantly longer (we use > to indicate these cases). In addition, for some analytes, stability information is presented in supplemental publications not included here.
There are multiple platforms for biomarker analyses, and labs vary in capabilities according to their investment in specific analytic systems and technologies. We note, in general terms, the analytic methods applied to each analyte since this may be a limiting factor for some labs.
The precision of an assay can be estimated by calculating the coefficient of variation (CV; standard deviation/mean) of multiple determinations of a single sample, all measured in a single batch. This is typically done with multiple samples across the full range of measurable values, but for ease of presentation, and since investigators differ in the number of samples they use to determine precision, we present the simple average intra-assay CV for each method. It is important to note, however, that the precision of an assay may vary across the assay range, and precision is often poorer at lower concentrations.
The day-to-day variation, or reliability, of a method can be estimated by calculating the CV of multiple determinations of a single sample measured on different days. As with precision, we present the average inter-assay CV as an approximation of assay reliability.
Lower detection limit
Sometimes referred to as analytical sensitivity, the lower detection limit of an assay is the smallest concentration of analyte that can be differentiated from zero with confidence. This is typically defined as the quantity of analyte that corresponds to a signal that is two or three standard deviations above the mean signal derived from multiple determinations of a sample free of analyte.
Blood spot/plasma comparison
The comparison of blood spot assay results with those from matched, simultaneously collected serum or plasma samples using a previously established, “gold standard” method is an excellent validation tool. Statistical evaluation of this relationship is typically performed with linear regression, or by inspecting residual plots for evidence of bias or inconsistent variability across the range of measurement. Analysis of matched blood spot and plasma/serum samples can also be used to generate a conversion formula to derive plasma-equivalent values from results with blood spot samples, although caution should be used in the application of plasma equivalents, since the relationship will vary across analytic methods, and may vary across populations.
Is the blood spot method presented in sufficient detail that a lab with appropriate analytic capabilities could reasonably expect to implement the method with success? We answer “no” if key information is missing that would require investigators to contact the method’s developers, or implement additional assay development steps prior to application.
Are all the materials required for the assay commercially available, or were key reagents (e.g., antibodies, calibrators) developed in-house? We answer “yes” if all reagents could be purchased from established suppliers at the time of publication. This is subject to change, as in-house reagents (or acceptable substitutes) may become available over time, and investigators are often generous in sharing their reagents. Conversely, previously available reagents for older methods may be difficult to obtain.