Study Design

One SR,¹³ two non-randomized studies,^23,24 and nine evidence-based guidelines^25–33 regarding folate testing were identified. The single SR,¹³ concerned the clinical utility of folate testing and searched evidence published from 2002 to 2013. No list of included studies was provided but four guidelines and two retrospective studies were summarized in varying degrees of detail. It included a review of guidelines as well as clinical and cost evidence. The two non-randomized studies^23,24, published in 2013²³ and 2015,²⁴ were retrospective chart reviews carried out at large medical institutions that cover the clinical utility of folate testing. One non-randomized study²³ was described in the SR¹³ but due to inadequate reporting it was also included separately in this review. Two evidence-based guidelines^28,31 included recommendations regarding which assay to use for folate testing, and all nine included recommendations regarding clinical indications for folate testing.^25–33

Country of Origin

The single SR was conducted by authors in Australia.¹³ The two non-randomized studies were conducted in the United States (US).^23,24 The evidence-based guidelines were developed by researchers in the United Kingdom,²⁸ US,^27,32 multiple European countries,^25,33 South Korea,²⁹ multinational locations,³⁰ and Canada,²⁶ though the Canadian guideline was adapted from a guideline originally developed in the US. All guidelines were published between 2010 and 2014.

Patient Population

The SR¹³ concerned patients at risk for folate deficiency including both apparently healthy patients and those with chronic diseases linked to folate deficiency.¹³ The two non-randomized studies included patients who underwent folate testing at large medical centers.^23,24 One study only included adults while the other did not specify age but did report that both inpatients and emergency department patients were included.²³ One study²³ specified indications for folate testing including hematological, neurocognitive, and dietary risk factors. The other study²⁴ did not specify indication. The evidence-based guidelines were geared towards patients at risk of folate deficiency,²⁸ with non-myeloid malignancies,²⁶ with Alzheimer’s and non-Alzheimer’s dementia,^25,29,31,33 with chronic kidney disease,³⁰ and bariatric surgery patients.^27,32

Interventions and Comparators

All of the identified references evaluated folate testing (either serum, RBC folate, or both). Only the SR¹³ specified comparators, including supplementation without testing, no testing, or in the case of comparing quality of testing, alternative folate assays.¹³

Outcomes

The SR¹³ and both non-randomized studies^23,24 investigated the clinical utility of folate testing. Clinical utility outcomes of interest included improvement in health outcomes and evidence of change in medical management following folate testing. The SR¹³ also investigated the comparative quality of various folate assays.

All of the evidence-based guidelines provided recommendations regarding the appropriate clinical indications for folate testing.^25–33 Several of the guidelines also discuss which folate assay (serum or RBC folate) is more appropriate for evaluating folate status.^28,31

Systematic Reviews

The SR¹³ set clear objectives but did not cite a published protocol. A comprehensive literature search was conducted on multiple databases as well as a grey literature search. However, the search was restricted to publications from 2002 onward and by language (English). No justification was given for the restricted search dates, so it is unclear whether potentially relevant evidence was not considered in formulation of results, and also reduces transparency due to the absence of clarity surrounding evidence synthesis. The primary concern was that no explicit list of included studies and study characteristics was provided. In general the review did not follow standard SR reporting format as set out by the Preferred Reporting Items for Systematic Reviews and Meta-Analyses Statement. Scientific quality was assessed with the National Health and Medical Research Council Dimensions of Evidence tool, and quality was taken into consideration in the formulation of conclusions. Quality of individual studies was not discussed explicitly. No pooling of results or assessment of publication bias was conducted. The review included disclosure of professional affiliations and funding sources none of which appeared to be industry-associated.

Non-Randomized Studies

The two non-randomized studies^23,24 were both based on health records; therefore, while they suffered from the typical limitations of retrospective studies including lack of randomization and blinding and possible misclassification bias, some design elements such as patient selection and follow-up did not pose a significant risk of bias. In terms of reporting, both studies stated their objectives, intervention, and study findings clearly. One study reported estimates of random variability and probability values,²⁴ but the other study did not as no group comparisons were conducted.²³ Neither study reported safety outcomes associated with folate testing. In addition, one study did not report demographic characteristics thoroughly.²⁴ As these were database linked studies, the external validity was reasonable for hospital patients in developed countries with folate fortification. One study conducted a thorough review of indications for folate testing in a subset of the total population but did not report demographic characteristics of this subgroup leading to unclear generalizability of these outcomes.²³ Neither study had the presence of design elements necessary to blind participants or assessors but the statistical analysis was appropriate and the main outcomes were measured appropriately. As both studies relied on chart data it is possible that a change in management may not have been recorded. Potential confounders were not considered in one study²³ and minimally assessed (i.e., age, body mass index, hematological profile, other micronutrient status, race, alcohol intake, use of supplements) in the context of group comparisons in the other, but did not include clinical conditions associated with folate deficiency.²⁴ Neither study disclosed a sample size calculation.

Evidence-Based Guidelines

There was substantial variation in the quality of evidence-based guidelines. Most guidelines had clearly described objectives,^25–33 target populations,^25–33 and explicitly stated health questions supporting their scope and purpose.^{25,26,26–32} The clinical expertise represented on the guideline development groups and the target users were clearly stated by most guidelines,^{25–27,29–33} but not all,^25,28,33 and included representation from multiple clinical specialties as well as methodological support. Specific health questions were not stated by one guideline.³³ Patient input was not sought by any of the guideline development teams but one guideline stated that patient preferences were considered when “a recommendation involves a substantial element of personal choice or values.”³⁰ The level of rigour of development varied. While comprehensive search strategies were employed by all guideline development teams, it was unclear whether the search was truly systematic in several cases as the number of reviewers involved in screening, selection, and abstraction tasks was not disclosed.^{25,27,28,32,33} Likewise, several guidelines failed to disclose their study selection methods^27,28,32,33 or disclose the number of reviewers involved in information synthesis.^27,28,32 Despite these drawbacks, the quality of evidence,^25–32 methods for formulating recommendations,^25–31,33 as well as consideration of risk-benefit profile, ^25–33 linkage between recommendations and supporting evidence,^26–28,31 and a peer-review process^25–31,33 were present in the majority of guidelines. One guideline failed to declare a peer-review process,³² and the direct linkage between evidence and recommendations was unclear in several.^25,32,33 In several cases it was clear that evidence search results did not include information to inform recommendations regarding folate testing, which relied on expert opinion.^29,30 Some guidelines^{27,29–31,33} disclosed a plan and timeline for updating the review, though dates were unclear in several cases. The clarity of presentation of recommendations was good in most cases.^{25,26,28–33} Key recommendations were embedded within text and not explicitly stated in some cases.^25,32,33 The area in which most^{25,26,26–28,32,33} but not all^29–31 guidelines were lacking was applicability. Facilitators and barriers to implementation, implementation tools, and a method of monitoring or auditing impact were not described. In cases where these elements are missing the potential and measured impact of these guidelines on practice and the culture of folate testing is difficult to gauge. Also, the lack of patient input may have had consequences for the scope of the guidelines and relevance to the needs of patient populations at risk for folate deficiency. All guidelines disclosed funding sources and competing interests and while it was unclear whether views of guideline group members influenced recommendations (especially in the case of clinical consensus) no funding sources were of great concern (i.e., no manufacturers or purveyors of folate testing).

What is the diagnostic accuracy of red blood cell folate versus serum folate testing for identifying folate deficiency?

No literature was identified regarding the diagnostic accuracy of RBC folate versus serum folate testing for identifying folate deficiency; therefore, no summary can be provided.

What is the comparative clinical utility of red blood cell folate versus serum folate testing?

No literature was identified regarding the comparative clinical utility of RBC folate versus serum folate testing; therefore, no summary can be provided.

What is the clinical utility of folate testing?

One systematic review,¹³ and two non-randomized studies^23,24 directly investigated the clinical utility of folate testing.

The SR¹³ did not identify any prospective studies regarding the clinical utility of folate testing. Two retrospective studies included in the review did not provide evidence to support the clinical utility of folate testing in patients with anaemia or dementia, and inpatients and emergency department patients, based on lack of change in management.¹³

Both non-randomized studies^23,24 failed to observe a meaningful change in medical management of patients following deficient folate tests. One study²⁴ reported that 39 to 56% of patients with deficient folate results received replacement therapy with no explanation as to why some patients did not. The other study only identified 2 (0.1%) deficient folate tests and 7 (0.3%) low normal folate tests, overall. Neither deficient patient received a subsequent change in management.

What is the cost-effectiveness of folate testing?

No literature was identified regarding the cost-effectiveness of folate testing; therefore, no summary can be provided.

What are the evidence-based guidelines regarding appropriate indications for folate testing?

The SR¹³ identified several guidelines suggesting that folate testing is indicated in patients with dementia, chronic fatigue syndrome/myalgic encephalomyelitis, patients with abnormal blood counts and those with suspected gastrointestinal disorders associated with malabsorption or malnutrition.

The evidence-based guidelines also recommend folate testing for patients with Alzheimer’s disease^29,31,33 and other non-Alzheimer’s dementia syndromes,^25,29,31 non-myeloid cancer patients at risk of anemia,²⁶ and CKD.³⁰

Recommendations based on poor quality and low subjective factor impact evidence (based on study design, data analysis, and interpretation factors)³² and weak recommendations based on low quality evidence²⁷ were made regarding bariatric surgery patients. One guideline recommended pre-partum folate testing for bariatric surgery patients who become pregnant post-surgery.³² Folate testing was recommended for patients who screen negative for iron deficiency anemia during routine evaluation post-surgery.³² Post-operative folate testing was also recommended for patients who undergo Roux-en-Y,³² laparoscopic biliopancreatic diversion with or without duodenal switch,^27,32 gastric bypass,²⁷ and sleeve gastrectomy procedures.²⁷

One guideline recommended that folate status be assessed in situations similar to vitamin B12 deficiency.²⁸ Based on the clinical indications listed this includes patients with anaemia, eating disorders, autoimmune diseases, history of glossitis or mouth ulceration, history of peripheral neuropathy, poor proprioception, malabsorption syndromes, use of certain medications (e.g., metformin, proton pump inhibitors, and oral contraceptives), neurocognitive impairment, consuming diets low in animal sourced foods, and pregnant women.

What are the evidence-based guidelines regarding which assay to use for folate testing?

Two evidence-based guidelines were identified regarding the optimal analytical method for folate testing.^28,30 The UK guideline²⁸ reported that serum folate testing is sufficient unless there is a strong clinical suspicion of folate deficiency in spite of normal serum folate and vitamin B12 levels. The Kidney Disease Improving Global Outcomes guideline³⁰ recommended that serum folate be used for assessment in most cases unless serum levels are normal or if recent dietary intake could potentially influence results.

Clinical Evidence

All of the direct evidence identified on clinical utility of folate testing was conducted in hospital settings. This limits generalizability to other clinical populations such as long-term care and community dwelling patients with conditions associated with folate deficiency. Thus, while clinical utility was not demonstrated in hospital patients in a folic acid fortified country, these results cannot be extrapolated to non-fortified regions and alternative populations with higher deficiency rates or risk.

There was discordance in the method of classifying folate deficiency (one study used a single cut-off and another used multiple) among the two non-randomized studies.^23,24 The cut-offs used to define folate deficiency have an impact on the rate of deficiency and consequently on the composition of the deficient population assessed for benefits of folate testing. This issue could have resulted in either over or underestimation of clinical utility. Future investigation into this cut-off disagreement and standardization is warranted as there is a lack of consensus on the topic.²⁴

Due to the retrospective design of the non-randomized studies,^23,24 the temporality of testing and medical management (e.g., supplementation, clinical intervention) was unclear. This problem arises in the results of the Theisen-Toupal et al. study²³ in which one patient with deficiency was shown to be receiving supplementation prior to testing. It was suggested that the patient was at risk for folate deficiency, not due to lack of intervention, but potentially for other reasons (e.g., lack of compliance).²³ Inclusion of these patients in analysis may skew the results towards a null effect.

Guidelines

The clinical expertise of the guideline development groups was unclear in several cases. As many recommendations were based on expert consensus the credibility of recommendations developed by these groups is unclear.

The large majority of the guidelines did not include recommendations regarding folate testing informed by high quality evidence. In fact, many recommendations were based on clinical consensus or expert opinion alone. In some cases this may have been due to lacking evidence, but in other cases limitations of the search may have resulted in relevant evidence not being considered. In addition, poor reporting of systematic review methodology, research questions and search and study selection methods limited transparency and reproducibility of several guidelines. There were also cases of unclear linkages between evidence and recommendations, which may detract from the credibility of some recommendations.

The indications embedded within the recommendations for folate testing from the evidence-based guidelines cannot be considered an exhaustive list. The absence of guidance on other populations at risk for folate related health outcomes (e.g., mothers with offspring at high risk of congenital malformations, methotrexate users) does not preclude them from being valid populations for folate testing; rather, there is no evidence-based guidance on these topics.

The guideline²⁸ that recommended folate testing for conditions similar to those tested for vitamin B12 deficiency did not discuss why folate testing was relevant for indications clearly only associated with vitamin B12 deficiency, such as pernicious anemia. The appropriateness of this general recommendation is unclear due to the lack of clarification, despite the strong recommendation being based on high quality evidence. It may relate to the potential for folate supplementation to mask neurological sequelae of vitamin B12 deficiency but this is unclear.

It should be noted that one guideline²⁷ used the terms folic acid and folate interchangeably. Folic acid refers to the synthetic form of folate that is used in many supplements and fortified products due to its enhanced stability and absorption. However, serum and RBC folate tests measure many folate vitamers; therefore, referring to folate testing as a folic acid test is inaccurate. Specific tests for folic acid levels do exist but are not common outside of the research setting.