Validation of Non-Invasive Hemoglobin Measurement Using Optical Method for Anemia Screening Among Adults in Jakarta (SPECTRO-Hb)

Background

Anemia affects billions of people worldwide and remains an important public health challenge, particularly in low- and middle-income countries. Undiagnosed anemia is associated with impaired physical performance, reduced productivity, cognitive dysfunction, poor maternal outcomes, and increased morbidity. Early detection of anemia is therefore essential for timely intervention and prevention of complications.

Conventional hemoglobin measurement relies on venous blood analysis using automated hematology analyzers or capillary blood testing using point-of-care devices. While these methods provide reliable results, they require blood collection, trained healthcare personnel, consumable materials, and appropriate laboratory infrastructure. In addition, invasive sampling may cause pain, anxiety, and reluctance to participate in large-scale screening programs.

Recent advances in optical and spectrophotometric technologies have enabled the development of non-invasive methods for estimating hemoglobin concentration without blood collection. These systems utilize light absorption and reflection characteristics of peripheral tissues to estimate hemoglobin levels rapidly and painlessly. Non-invasive approaches may improve participant acceptance, facilitate repeated measurements, and support large-scale community screening initiatives. However, their accuracy and agreement with established laboratory methods remain variable across populations and settings. Therefore, local validation studies are necessary before widespread implementation.

Objectives

Primary Objective

To evaluate the validity and agreement of a non-invasive hemoglobin measurement method compared with capillary POCT hemoglobin measurement and venous blood hemoglobin measured using an automated hematology analyzer.

Secondary Objectives

To determine the correlation between non-invasive hemoglobin measurements and hematology analyzer results.

To assess agreement between non-invasive measurements, POCT results, and hematology analyzer measurements using Bland-Altman analysis.

To evaluate the sensitivity, specificity, positive predictive value, negative predictive value, likelihood ratios, and diagnostic accuracy for anemia detection.

To explore the feasibility of implementing non-invasive hemoglobin measurement in community-based screening programs.

Study Design

This study is a community-based cross-sectional diagnostic validation study.

Study Population

The study population consists of adults aged 18 years and older participating in community health screening activities in Kelurahan Kota Bambu, Jakarta.

Sample Size

A total of 150 participants will be recruited using consecutive sampling. Sample size calculation was based on an expected diagnostic accuracy of 90%, a confidence level of 95%, and a margin of error of 5%, with additional participants included to compensate for incomplete data.

Study Procedures

Each participant will undergo three hemoglobin measurements during a single study visit:

Non-Invasive Hemoglobin Measurement Performed on the thumb using a using optical method device. No blood collection is required. Capillary Hemoglobin Measurement Obtained through finger-prick blood sampling. Measured using a point-of-care testing (POCT) device. Venous Hemoglobin Measurement Venous blood collected by trained healthcare personnel. Analyzed using an automated hematology analyzer as the reference standard.

Demographic information, medical history, smoking status, anthropometric measurements, and factors potentially affecting measurement accuracy will also be collected.

Outcome Measures

Primary Outcome

Agreement between non-invasive hemoglobin measurements and venous hemoglobin measured using a hematology analyzer.

Secondary Outcomes

Correlation coefficients between methods. Mean measurement bias and limits of agreement. Sensitivity and specificity for anemia detection. Positive and negative predictive values. Likelihood ratios. Area under the receiver operating characteristic (ROC) curve. Cohen's kappa coefficient for anemia classification agreement. Statistical Analysis

Descriptive statistics will summarize participant characteristics and hemoglobin measurements. Agreement between methods will be assessed using Bland-Altman analysis and Intraclass Correlation Coefficients (ICC). Correlation analyses will be conducted using Pearson or Spearman tests as appropriate. Diagnostic performance will be evaluated through sensitivity, specificity, predictive values, likelihood ratios, ROC analysis, and area under the curve (AUC). Cohen's kappa coefficient will assess agreement in anemia classification. Subgroup analyses may be performed according to age, sex, and measurement conditions.

Expected Impact

The study is expected to provide evidence regarding the accuracy and screening capability of non-invasive hemoglobin measurement in community settings. Results may contribute to the development of accessible, rapid, and patient-friendly anemia screening programs and support future research involving pediatric and other vulnerable populations.