Karen Titus
The distance between recognizing the issues of iron deficiency and related anemia, and providing a diagnosis and treatment, can often stretch months to years. Improvements in the field have been more tortoise than hare.
But the pace may be picking up. This fall, the American Society of Hematology published guidelines for the diagnosis of iron deficiency (Powers JM, et al. Blood Adv. 2026;10[21]:7353–7374). Key recommendations include ferritin thresholds for specific populations, aimed at diagnosing iron deficiency regardless of the presence of anemia.
The new guidelines set forth five recommendations:
- In children ages nine months to four years, the guideline panel suggests using a serum ferritin threshold of ≤20 ng/mL instead of a threshold of ≤12 ng/mL for diagnosing iron deficiency.
- In the adult population (excluding menstruating or pregnant individuals), the panel suggests using a serum ferritin threshold of ≤30 ng/mL instead of ≤15 ng/mL for diagnosing iron deficiency.
For those with signs or symptoms associated with iron deficiency or ongoing risk factors, they note, a threshold of ≤50 ng/mL is appropriate for diagnosing iron deficiency and guiding management decisions.
- In menstruating individuals, the panel suggests using a serum ferritin threshold of ≤30 ng/mL and suggests against using a threshold of ≤15 ng/mL for diagnosing iron deficiency.
All three of these, the authors note, are conditional recommendations based on low certainty in the evidence about effects.
- In pregnant individuals, the panel suggests using a serum ferritin threshold of ≤30 ng/mL and recommends against using a threshold of ≤15 ng/mL for diagnosing iron deficiency.
The first suggestion is a conditional recommendation; the latter is a strong recommendation.
For pregnant individuals with anemia, with or without other iron deficiency risk factors and/or associated symptoms, they further note, a serum ferritin threshold of ≤50 ng/mL may be appropriate to diagnose iron deficiency and guide management decisions.
- For adults with anemia of inflammation, the panel suggests performing both serum ferritin and transferrin saturation, rather than serum ferritin alone, to evaluate for iron deficiency.
This last recommendation is also conditional, based on very low certainty in the evidence.
The guidelines note that many laboratories use a laboratory-defined serum ferritin lower limit of normal as opposed to a clinical decision limit. The concern, the authors write, is that using a LLN results in a significant variation in reference ranges between labs; moreover, it classifies values that are consistent with iron deficiency within the reference range. Both contribute to underrecognition of iron deficiency by ordering providers. Using a clinical decision limit, on the other hand, would let laboratories flag results at or below a clinically relevant diagnostic threshold, regardless of the reference range.
The guidelines also acknowledge the wide variability across testing methods and lack of harmonization across platforms, both of which have long bedeviled the field.
Because it’s not yet a standardized assay, clinicians may not realize ferritin results are method dependent so values can differ among laboratories. “These interassay differences must be considered when interpreting patient data using a single clinical decision threshold listed in a clinical guideline,” says Jaime Noguez, PhD, D(ABCC), senior attending in pathology; section head of clinical chemistry, diagnostic immunology, and biochemical genetics; and director of chemistry and toxicology, University Hospitals Cleveland Medical Center. That makes it especially important, she says, for laboratories to understand the performance of their specific ferritin assay and communicate these limitations when implementing the new clinical decision thresholds.
Laboratories should also bear in mind that the ferritin reference intervals are not the same as the clinical decision thresholds, and it’s necessary to distinguish one from the other. “Laboratories can play an important role in helping clinicians understand how these thresholds should be interpreted and applied,” says Dr. Noguez, who is also associate professor of pathology, Case Western Reserve University School of Medicine, and a member of the CAP Clinical Chemistry Committee.
As laboratories consider implementing the new guidelines, they’ll need to consider how ferritin results should be presented in the laboratory report and electronic health record. One approach is to retain the laboratory’s established population-based reference interval and add a comment that provides the ASH clinical decision thresholds and appropriate clinical context as interpretive guidance. “This preserves the distinction between a reference interval and a clinical decision limit,” Dr. Noguez says, “but it also creates the possibility that a result may appear ‘normal’ based on the reference interval while meeting a guideline-based threshold for iron deficiency.”
An alternative approach is to replace the lower reference limit with the ASH clinical decision limit, which may provide clinicians with a clearer signal for identifying potential iron deficiency, she says, but will no longer represent the conventional population-based reference interval.
“There are reasonable arguments for both approaches,” Dr. Noguez says. “Reference intervals and clinical decision limits serve different purposes, so laboratories need to be deliberate about how each is presented in the medical record.” Whichever strategy a laboratory chooses, she says, the goal should be to make the distinction sufficiently clear “that clinicians can appropriately interpret and act on the result.”
This is not a new concept for laboratories, Dr. Noguez says. “For many analytes, clinical interpretation depends on decision limits established from clinical evidence rather than population-based reference intervals alone, and ferritin may increasingly need to be approached in that same way.” The laboratory can help bridge that gap, she adds, by providing clinicians with the context they need to interpret the result appropriately.
All of which means laboratories will have their work cut out for them regardless of whether they decide to adopt the new guidelines. If nothing else, the guidelines are bound to raise questions from clinicians and perhaps nudge their laboratory colleagues to reconsider current practices.
The guidelines fill a gap, however imperfectly, observers say. The authors note that iron deficiency is the most common micronutrient deficiency and that it progresses to anemia when left untreated.
Dr. Noguez calls the guidelines “a good start” in tackling the problem. “It’s one iteration of many.” It’s also an improvement over what physicians have relied on, particularly in that it addresses the longstanding variability in testing practices for iron deficiency.
As the guidelines’ authors explain, variability across methods means different assays can produce different serum ferritin values on the same sample. And while this occurs across a range of values, it “is most pronounced, and potentially clinically significant, at the higher end of serum ferritin values,” the authors write.
“This is an extremely important topic because it’s so ubiquitous,” says Darryl Palmer-Toy, MD, PhD, physician director of SCPMG Regional Reference Core Laboratories, Kaiser Permanente–Southern California Permanente Medical Group. “Our primary care providers are wrestling with questions of anemia, iron deficiency, and iron deficiency anemia on a daily basis.” One problem, of course, is that the symptoms of iron deficiency—fatigue, cognitive impairment, shortness of breath—mimic those seen in other conditions. Ferritin is an acute phase reactant. And while there are many markers, none of them are ideal. “All of them have significant limitations.
“So it’s really important that ASH focused attention on what might seem to be a fairly mundane topic,” he continues.
But he’s concerned about the limitations that the guideline authors themselves identified—that there remains a dearth of strong evidence in the field. “Most of the guidelines had only low levels of confidence based on the available evidence,” says Dr. Palmer-Toy, who is a member of the CAP Accuracy-Based Programs Committee. “So clearly more work needs to be done.”
Speakers in a session at the Association for Diagnostics and Laboratory Medicine annual meeting in July addressed the new thresholds and their experience with adopting them. Guideline coauthor Michelle Sholzberg, MDCM, MSc, cited data from the National Health and Nutrition Examination Survey (Weyand AC, et al. JAMA. 2023;329[24]:2191–2193) in noting that in nonpregnant females between ages 12 and 21 years, the prevalence of iron deficiency using a serum ferritin threshold of less than 25 ng/mL was 38.6 percent; the prevalence of iron deficiency anemia was 6.3 percent. “It’s common,” said Dr. Sholzberg, a hematologist and medical director of the coagulation laboratory and associate professor and clinical scientist, St. Michael’s Hospital and Li Ka Shing Knowledge Institute, University of Toronto, Ontario.
Data from Canada showed similar findings in nonpregnant females ages 15 to 54 (Wen S, et al. Am J Hematol. 2024;99[8]:1492–1499). Prevalence of nonanemic iron deficiency was 40 percent, while prevalence of iron deficiency anemia was 13 percent. Most alarming, Dr. Sholzberg said, was that 56 percent of those in the latter group had a normal mean corpuscular volume. “The CBC, therefore, is not an effective screening test for iron deficiency.”
WHO guidelines have not adequately addressed the problem, Dr. Sholzberg and others contend. Those guidelines “suggest that a serum ferritin threshold of less than 12 or 15 is in keeping with iron deficiency,” Dr. Sholzberg said, despite data that show that serum ferritin reference interval establishment studies don’t exclude those with iron deficiency risk factors in the apparently healthy reference population. Using a reference population that includes those with iron deficiency “is obviously very problematic.”
Another presenter, Dana Nyholt, PhD, D(ABCC), clinical biochemist, LifeLabs (a subsidiary of Quest Diagnostics), Canada, urged attendees to adopt the clinical decision limits but acknowledged laboratories’ concerns about the move.
She called the guidelines “a call to action for physicians for further investigation, but it’s not a black-and-white diagnosis.” She also acknowledged the concern that a CDL of 30 ng/mL will cause too much flagging in the laboratory. Among other worries: Ferritin is context-dependent and affected by factors such as inflammation; moreover, it might be overly confusing for clinicians to communicate a low-end clinical threshold without an upper-end threshold for iron overload.
The guidelines “are a guideline to act, but they are not a substitute for clinical judgment,” Dr. Nyholt said. At LifeLabs, which implemented the ASH guidelines several years ago, clinical education was “the most critical” step in the process. “We spent a lot of time educating clinicians on what the clinical decision limit was, how it would impact their clinical management,” and cautioning them that their flagging rates would likely increase significantly.
“We braced for clinical complaints,” Dr. Nyholt said, given that, as a large reference center for outpatient testing, the laboratory runs millions of ferritin tests annually. And how many did they receive? Not a single one, she said.
In the Cleveland area, University Hospitals has already adopted the 30 ng/mL threshold across its obstetrical network, spurred by both its OB Network and the laboratory. The UH guideline has been in place for at least five years, according to Nancy Cossler, MD, chief of system quality for obstetrics, University Hospitals, and vice chair of quality and patient safety, MacDonald Women’s Hospital. UH plans to implement the ASH guidelines more widely across the system, and their experience in managing anemia peripartum might offer a glimpse into what a wider rollout might entail.
Noting that the ASH guidelines call out pregnant patients in the fourth recommendation, Dr. Noguez says, “I think our obstetric team”—with Dr. Cossler as the clinical lead for the initiative—“was ahead of the curve on this. They recognized the issue of underdiagnosis of anemia in pregnancy early on and, through close collaboration with the laboratory, implemented a more proactive approach to identifying iron deficiency.” At their first prenatal visit, all patients receive a complete blood count anemia panel, with reflex testing for ferritin as well as folate, vitamin B12, total iron-binding capacity, and iron if indicated.
“We screen them every trimester for iron deficiency,” Dr. Noguez says, “as well as other necessary laboratory testing. We’ve been using a ferritin threshold of 30 ng/mL even before the new ASH guidelines were released, recognizing that using a higher threshold allows us to identify iron deficiency earlier, before it progresses to anemia.” They recognize this approach may not identify every case of iron deficiency during pregnancy, but “we believe it represents a meaningful step toward earlier detection and a more proactive approach than what has historically been used.”
That emphasis on earlier identification is an important strength of the ASH guidelines, she says. “The higher thresholds move the field toward earlier recognition of iron deficiency, rather than waiting for patients to become symptomatic or anemic.” Dr. Noguez also lauds the guidelines’ focus on other populations at increased risk of developing iron deficiency, including patients who are menstruating and children. “At the end of the day, this is about improving how we recognize iron deficiency so that all patients have the opportunity to receive timely and appropriate care.”
For pregnant patients, the 30 ng/mL cutoff has altered care considerably, says Dr. Cossler, who is also associate professor, Department of Reproductive Biology, Case Western Reserve University School of Medicine. “The work we did with the lab in trying to maximize antepartum hemoglobin levels is based on the fact that when we looked at our blood transfusions at UH,” a minority but still substantial number—perhaps 15 percent of blood transfusions, she estimates—occurred because women arrived for delivery with a low hemoglobin. They then experienced the usual blood loss associated with vaginal delivery. “It’s always been underquoted,” Dr. Cossler says. “People often say it’s 500 mL but it’s really close to 1,000.” For Cesarean deliveries, which account for 28 to 30 percent of deliveries across the system, blood loss ranges from 1,000 mL to 1,500 mL, she says.
Dr. Cossler continues: “If a woman starts with a hemoglobin of 8, she can’t tolerate that blood loss.” The goal was “to protect women from transfusion—and there was that chunk of women who were getting transfused because we hadn’t maximized their hemoglobin.” The UH physicians were racing against the clock. “You only have a short period of time to do that prior to delivery. We have a very robust guideline to help us boost hemoglobin.” In addition to helping the mother, the pediatric and obstetric literature has shown that chronic anemia likely has detrimental effects on fetal neural development, she adds.
Adopting the 30 ng/mL threshold for pregnant individuals was clinician-driven in one respect, Dr. Noguez says—they were clinicians who identified symptoms despite blood work in the normal range. But the laboratory also had to consider how these results are presented in the medical record so that clinicians can readily recognize when a result meets a decision threshold and determine the appropriate next step, Dr. Noguez says.
“Successful implementation involves much more than just changing a cutoff,” she says. “The laboratory and clinical teams need to consider reporting, interpretation, clinical education, and follow-up so the guideline can be applied appropriately in practice. So it’s truly a team effort, one largely driven by clinicians based on their experience—the outcomes and symptoms they see in their patients—but definitely a team effort.”
The UH guideline includes reflex testing, “which the lab helped us build,” Dr. Cossler says. “If this is hemoglobin, and it’s microcytic, what do we need to know to illuminate other causes?”
“The lab was great, and they helped us build those tests automatically, so people don’t miss them,” Dr. Cossler reiterates.
(She doesn’t take that for granted. When it came to managing outpatient laboratory testing, performed by a reference laboratory, “we had to fight with them,” she says with a laugh. “They were trying to tell us the normal values for hemoglobin, and not taking into account whether a patient was iron deficient. And don’t tell us what the normal values are in a pregnant woman—we know what the normal values are. We know what we’re aiming for.” In contrast, she says, the UH laboratory colleagues “were a huge help.”
Dr. Cossler adds: “There’s a widespread failure to understand that even if the hemoglobin is relatively normal, it doesn’t mean the patient’s not iron deficient.” She attributes that partly to the fact that “It’s hard to get people to include pregnant women in studies. It’s understudied and underappreciated.”)
How well has the cutoff of 30 ng/mL worked for pregnant patients at University Hospitals?
“It’s a hard question to answer,” Dr. Noguez says, “since we are still collecting and analyzing outcomes data. But our preliminary findings are encouraging and suggest we are moving in the right direction.”
By incorporating ferritin testing with this threshold into trimester-specific order sets, University Hospitals standardized screening practices across clinicians. “Adding ferritin to our order sets has helped ensure more consistent screening,” Dr. Noguez says, which has made it possible to identify more patients with iron deficiency and initiate treatment promptly, ultimately contributing to higher hemoglobin levels when they are admitted for labor.
But in some ways, Dr. Noguez says, she and her UH colleagues face challenges similar to those encountered by the creators of the ASH guidelines. “One of the primary limitations is the lack of robust prospective data to inform clinical recommendations. This is a challenge shared by many other diagnostic guidelines, where decisions must be made based on the best available evidence, even when that evidence is incomplete. The available evidence isn’t perfect.”
“Outcomes data is incredibly hard to generate.”
Indeed, she adds, experts rely primarily on observational studies, or those that use comparisons against imperfect reference standards, or that evaluate diagnostic performance of an assay rather than direct patient outcomes.
Her hope is that once users become familiar with the new guidelines, and see the limitations of available evidence, it will be “a call to action in the field: This is the type of data we need. We need to capture what we’re seeing. We need to publish it, so we can throw more evidence behind these recommendations. This opens the community’s eyes to what data is needed.”
Dr. Cossler reports that she and her OB colleagues made at least one eye-opening discovery. Though it’s hardly news that blood supplies are tenuous, and that individuals who are delivering are at risk for hemorrhage, they were startled to find—when they looked more closely— the aforementioned group of women who were transfused because they came in with lower (8 ng/mL) hemoglobins at time of delivery. Though the physicians were aware of this anecdotally, “we didn’t know this was a low-hanging-fruit opportunity” to improve practices, Dr. Cossler says. “That was a surprise to us.”
Dr. Noguez is optimistic about the potential impact of the ASH guidelines, despite the limitations of the available evidence.
“I view these guidelines as an important step in the evolution of how we approach the identification and management of iron deficiency,” she says, “rather than a directive to immediately redesign existing laboratory workflows.” The ultimate goal is to identify iron deficiency earlier, before patients become symptomatic or progress to anemia.
Successful implementation, she says, requires a multidisciplinary approach tailored to each institution’s clinical and laboratory practices.
“Implementing these guidelines should be a collaborative effort involving clinicians, laboratory medicine specialists, and other key stakeholders to ensure the recommendations are appropriately integrated into clinical practice,” she says. And distinguishing between laboratory reference ranges and clinical decision thresholds is important. “Adopting these guidelines does not necessarily mean institutions need to change their existing reference ranges.”
When UH implements the guidelines more widely, Dr. Noguez says, their approach will likely be to maintain their established reference ranges while incorporating interpretive comments. “This will help clinicians understand and apply the updated recommendations appropriately, supporting more informed decision-making without significantly altering existing laboratory reporting practices.”
Clearly the OB Network was a knowledgeable clinician group. But as Dr. Noguez points out, educational needs may vary by clinical specialty and patient population. “The approach to interpreting iron studies may differ depending on whether a clinician primarily cares for otherwise healthy individuals, children, or patients with complex conditions such as cancer,” she says. “So laboratories need to recognize these differences and tailor their educational efforts, including providing guidance on when additional iron studies may be necessary.”
The guidelines identify situations in which transferrin saturation can help clarify ferritin results—for example, when a high ferritin might be due to underlying inflammation rather than adequate iron stores. “The appropriate diagnostic thresholds must be considered within the context of the patient’s underlying condition and clinical presentation,” Dr. Noguez says, noting the “considerable nuance” in interpreting these results and that a single threshold may not be equally applicable across all patient populations.
The more conservative threshold suggested in the guidelines may result in some patients being classified as iron deficient when they are not. “We recognize that increasing diagnostic sensitivity may come at the expense of specificity,” Dr. Noguez says. “However, the priority is to identify patients who might otherwise go undiagnosed and miss the opportunity for timely treatment.” When the potential consequences of untreated iron deficiency are weighed against the relatively low risks of appropriate iron supplementation, she adds, “a more sensitive approach can offer meaningful clinical benefit.”
Dr. Noguez draws a parallel to the introduction of high-sensitivity troponin assays.
“I think it’s a similar situation, though I know not everyone would agree. When those assays were first introduced, there were significant concerns that increased detection would lead to more positive results, unnecessary cardiology consultations, and additional demands on clinical workflows.”
“However, the increased sensitivity also enabled earlier identification of myocardial infarction, allowing more timely intervention and potentially reducing myocardial injury,” Dr. Noguez says. Many of the concerns about overdiagnoses and workflow interruptions didn’t bear out. “While these concerns are valid and deserve careful consideration, they should not become barriers to adopting diagnostic approaches that allow us to identify and treat patients whose conditions might otherwise go unrecognized.”
Dr. Palmer-Toy finds the guidelines to be more problematic.
On one hand, he praises the authors for being “fairly savvy compared to other guidelines groups in regard to laboratory-related issues.” Though he says it might have been helpful for the guidelines to consider mean cellular hemoglobin concentration as well as the admittedly more specialized reticulocyte hemoglobin concentration as complementary parameters, they avoided one common pitfall. “I think it’s good that they didn’t dwell on mean corpuscular volume [MCV], which is often the first question in many algorithms. MCV is a very limiting criterion in which to assess anemia.”
He also lauds their acknowledgment that not all ferritins are measured equally, noting that one study reported substantial differences around the clinical decision point (Braga F, et al. Clin Chem. 2022;68[9]:1202–1210). In fact, the Accuracy-Based Programs Committee discussed the ASH guidelines when it met this fall, he says, and is exploring ways to assess the harmonization of ferritin assays.
But he’s less sanguine about what he sees as a dismissive attitude toward soluble transferrin receptor, which is considered to be a less widely available assay. By focusing on transferrin saturation, he says, “you run into sensitivity problems. Just as ferritin rises with inflammation, TSAT drops with inflammation.” Soluble transferrin receptor, in contrast, is thought to be fairly insensitive to inflammation.
“That’s often a diagnostic dilemma,” Dr. Palmer-Toy says. Is the anemia attributable to chronic inflammation, or is it a consequence of iron deficiency? It’s not always an easy distinction to draw.
He raises other concerns as well.
The current WHO definition of anemia, which is used in the guideline, has the benefit of being widely adopted, Dr. Palmer-Toy says. “But I don’t think it’s the last word in defining anemia.” Rather than relying on an arguably one-size-fits-all approach, he says, “there’s actually a fair amount of literature that suggests there are significant differences across” countries, for example.
He also questions what he calls “the suspiciously round” clinical decision points (20, 30, 50, 100 ng/mL). He says that data from his own institution, as well as from Mayo Clinic Laboratories, for example, show slightly different cutoffs depending on sex. At Kaiser Permanente–Southern California, he says, there is a slightly higher normal range for men and a slightly lower one for women. A gender-neutral cutoff is not necessarily a well-supported assumption, “in my opinion.”
And while the issue of iron deficiency is important, he wonders if the new cutoffs might lead to overdiagnosis. Following up on the initial diagnosis could be “a budget-crushing consideration.” Access to follow-up procedures, such as endoscopy, could also be problematic. Moreover, he says, noncompliance is a notorious problem with iron supplementation.
He’s also concerned about a lack of specificity on the low end. “What defines low?” he asks. “And is that really even true?” He suspects there’s likely a spectrum of people who have different degrees of ferritin responsiveness to iron deficiency as well as to inflammation. “I can’t tell you how prevalent that is, but it does seem to happen,” he says. “It’s certainly not a rare thing to find people with ferritin very much lower than 30 who have no other evidence of iron deficiency.”
Ultimately, he says, “It’s clear that the evidence is still pretty thin, even by the grades they give their own recommendations. But I’m grateful for ASH for looking into this.”
Karen Titus is CAP TODAY contributing editor and co-managing editor.