Validation of a smartphone-based point-of-care hemoglobin assay for use in dogs.

OBJECTIVE

To validate a smartphone-based spectrophotometric assay for point-of-care (POC) measurement of hemoglobin (Hgb) concentration against the standard measurement method in dogs without increases in serum lipemia, hemolysis, or bilirubin.

DESIGN

Prospective observational study.

SETTING

University teaching hospital.

ANIMALS

One hundred thirty-nine dogs that had a CBC and corresponding biochemical profile submitted to the clinical pathology laboratory.

INTERVENTIONS

None.

MEASUREMENTS AND MAIN RESULTS

Blood tubes submitted for CBC were collected for hemoglobin (Hgb) measurements performed on a POC smartphone device. Each whole blood sample was run on the smartphone in duplicate using 2 strips, for a total of 4 POC Hgb readings per dog. Data collected for each dog included CBC Hgb value, 4 POC Hgb values, and select biochemistry values (bilirubin, lipemia, hemolysis, icterus). A calibration equation was estimated using a weighted linear regression: estimated CBC = (avgPOC-0.4871)/1.0015. For each dog, the percent error was computed between estimated and actual Hgb values; 95% of the percent errors ranged from -13.2% to 20.1%. The standard deviation of percent errors was 7.9% overall. When samples were further divided according to CBC Hgb concentration (low, normal or high), the standard deviation of percent error was 6.7% when Hgb<13 g/dL [130 g/L], 8.9% when Hgb 13-20 g/dL [130-200 g/L], and 6.5% when Hgb>20 g/dL [200 g/L]. The coefficient of variability among the 4 individual POC readings was 3.4%.

CONCLUSIONS

A smartphone-based photometric method for measuring Hgb represents a clinically useful POC alternative to a standard laboratory Hgb measurement. There was excellent intrasample reproducibility, and the standard deviation of percent errors was relatively constant across CBC ranges. Additional sampling of patients with a greater range of diseases and biochemical abnormalities that may influence spectrophotometric assays (ie, abnormal bilirubin, icterus, lipemia, hemolysis) is warranted to extend the findings of this study.