Diagnosis & Management of Vitamin B12 Deficiency in Primary Care – Are we following the guidelines?

Background: Vitamin B12 deficiency is common in primary care but its treatment practices vary across centres. One important cause of B12 deficiency is pernicious anaemia (identified by intrinsic factor antibodies), which is a risk factor for developing gastric carcinoma, and should be excluded. Additionally, recent evidence has suggested that some patients have been continued on B12 injections with no clear clinical indication. Recently,guidelines were produced to improve the investigation and management of B12 deficiency. Hence, this audit studied the investigation and management of B12 deficiency and adherence to clinical guidelines in a general practice (GP) in north-west England.Aim: To evaluate the appropriate diagnosis and management of vitamin B12 deficiency in our primary care centre against recognized standards.Methods: Clinical data of patients currently on oral cyanocobalamin and/or intramuscular hydroxocobalamin injections over a 1-year period were audited.Results: Thirty-eight patients (66% females, 34% males) receiving treatment for B12 deficiency were identified. Of these, 55% (21/38) had intrinsic factor antibodies checked and 52% (13/25) were managed according to the guidelines. 100% (8/8) of patients with dietary B12 deficiency (non-vegans) and 75% (3/4) of B12-deficient patients on long-term metformin have had follow-up serum B12 monitoring.Discussion: There is a need to improve the investigation for B12 deficiency, adherence to clinical guidelines, and documentation of patients’ diagnoses, treatment plans, dietary statuses, and required monitoring. We anticipate that adhering to guidelines when appropriate, with clear documentation, will improve the diagnosis and management of vitamin B12 deficiency so that safe prescribing and potential cost savings can be achieved.


Introduction
Vitamin B12 deficiency is commonly encountered in primary care, particularly among the elderly. 1 Although most clinicians are familiar with this condition, the approach to its diagnosis and treatment varies across practices, with the administration of oral, intramuscular, or other preparations of vitamin B12 being debated extensively. 2reover, it is controversial whether all patients who begin treatment for low serum B12 levels need to be on replacement therapy for life, as it has been argued that some patients continue to receive B12 replacement despite normal post-treatment serum B12 levels (defined as levels between 150-900 ng/L).Recently, the Manchester Anaemia Guide 3 was published, with reference to the NICE Clinical Knowledge Summaries (CKS), 4 to provide clinical guidance on these issues.It emphasised that, as part of the diagnostic process for every patient with suspected B12 deficiency, pernicious anaemia (PA), should be excluded by checking for intrinsic factor antibodies, as affected patients have an increased risk of developing gastric carcinoma. 5Moreover, PA is associated with many other autoimmune conditions (e.g.Graves' disease and vitiligo). 6ence, identifying these patients would inform the need for further work-up or referral to a specialist for evaluation in the presence of any significant clinical finding that suggests an underlying gastric carcinoma or autoimmune condition.In view of these issues, we conducted an audit to address them.

Aims 1. Assess whether patients with vitamin B12
deficiency have been appropriately diagnosed, managed, and followed up according to the Manchester Anaemia Guide at a GP practice in northwest England.

Methodology
The inclusion criteria used in the selection of patients were as follows: 1) All patients on vitamin B12 replacement (in the form of oral cyanocobalamin tablets or intramuscular hydroxocobalamin) between 1 April 2013 and 1 April 2014 were identified through our clinical database (Emis Web).2) From this group, clinical data was obtained (from Emis Web and Docman) and audited retrospectively.

Results
A total of 38 patients (66% female, 34% male) on B12 replacement were identified (figures 2 and 3), of which 4 (11%) presented with neurological symptoms associated with B12 deficiency (where other causes have been excluded) whereas the remaining 34 patients (89%) did not present with any neurological features (figure 2).The mean age of the patients is 62 years (range 23-93).Our patients with B12 deficiency-associated neurological symptoms were not referred for any specialist neurology evaluation as they were otherwise clinically stable.In patients with multiple causes, other contributing factors such as long-term use (>12 months) of proton pump inhibitors (PPIs) (e.g.omeprazole), previous gastric/colorectal surgery (e.g.gastric bypasses, Nissen's fundoplication), and terminal ileum carcinoma have been identified (figure 3C).
Results of Standard 1 (figure 4): IFA tests were performed in 21 patients (55%).Of these, 2 were positive for IFA whereas 17 were not, and the remaining 2 patients had an inconclusive result due to insufficient samples.Seventeen patients (45%) did not have an IFA test.All patients who had tested positive for IFA did not need to be referred to the appropriate specialists for further evaluation as they were clinically stable and responsive to vitamin B12 replacement.
Results of Standard 2 (figure 5): 13 patients (34%) were managed according to the guidelines (i.e. they were put on the most appropriate treatment regime based on the cause of their deficiency and the presence of neurological symptoms).Of these, 3 needed an initial IFA test and 1 needed a repeat test (despite an established cause) due to previously inconclusive results.Twelve patients (32%) were not managed according to the guidelines and the remaining 13 patients (34%) were excluded from the results due as the causes of their deficiencies were unascertained.Hence, after correction, 52% (13/25) of patients with an established cause of B12 deficiency had been managed according to guidelines.
Results of Standard 3a (figure 6): In total 9 patients had dietary B12 deficiency (in isolation or in combination with other factors).All 9 patients had their B12 levels corrected with replacement therapy between 1 April 2013 and 1 April 2014 (which was subsequently stopped) and with improved diet, after which follow-up monitoring was commenced.Of these, 8 patients (89%) had their follow-up B12 serum levels checked, but 4 of these 8 patients had the tests performed after the 6-month interval (to account for certain unavoidable factors that can result in a patient not being followed up within 6 months, we considered a follow-up blood test performed more than 1 year from the discontinuation of treatment to be late).One other patient (11%) was not yet due for follow-up and hence, was excluded from the results.In summary, 100% of patients (8/8) had their follow-up B12 levels monitored but only 50% (4/8) of patients had this within the appropriate time interval.
Results of Standard 3b (figure 7): In total, 8 patients had B12 deficiency as a result of long-term metformin use (in isolation or in combination with other factors).Of these, 3 patients (37.5%) had their follow-up B12 test done but 2 of them had the tests beyond the 6-month interval (>1 year late).Four patients were not yet due for follow-up (hence were excluded) and 1 remaining patient has not had any follow-up.In summary, 75% of patients (3/4) had their follow-up B12 levels monitored but only 25% (1/4) of patients had this within the appropriate time interval.Table 1 summarizes the results of the audit.

Discussion
The standards set were not fully met.There are several possible explanations for this.For one, it is possible that not all of the primary care physicians were aware of the new updates in guidance before the audit period, which may have dampened adherence to guidelines.Moreover, patients may not have been fully compliant with treatment, as the majority of our patients are receiving the B12 injections in a specialist nurse-based appointment, which may have been missed, hence disrupting treatment schedule.
A possible oversight of the clinician may result in IFA tests not being arranged.The test may also have been deemed unnecessary, as a definitive cause of patients' B12 deficiency may have already been identified on clinical evaluation.
With regard to follow-up monitoring, it is important to note that the 6-monthly monitoring interval for patients whose B12 deficiency is attributed to a dietary cause and/or long-term metformin use, as outlined in the Manchester Anaemia guide, serves only as a general guide.Currently, there is little evidence to guide monitoring in these groups, and regular monitoring has not been recommended in established guidelines, such as the NICE Clinical Knowledge Summaries. 4Again, clinicians who are not familiar with the Manchester Anaemia Guide would not be aware of the monitoring suggestions in the guide and hence, may not have monitored their patients as suggested.The lack of a documented reminder for a clinical or prescription review may have resulted in patients continuing on unnecessary B12 injections or being put on the incorrect treatment regime.
In conducting this audit, we noted some limitations.Firstly, there may be patients who have not been included in our study due to a personal choice not to undergo testing or receive any treatment for their B12 deficiency.Also, lapses in medical record keeping could have arisen as patients may have had testing conducted outside primary care that were not recorded or communicated to their primary care physician.Furthermore, not every patient could be allocated to a specific category for the management guidelines due to multiple or unascertained diagnoses.Hence, patients with an unascertained cause were excluded from tabulating the results of Standard 2. However, we will continue to monitor this group of patients.Lastly, the IFA test is highly specific (nearly 100%) but poorly sensitive (50-60%), [6][7] and this means that false-negative results could have occurred.In our study cohort, although 17 out of 19 patients tested negative for IFA, it is debatable whether a true diagnosis of PA can be excluded.
Furthermore, another important point to note is that vitamin B12 deficiency may arise in the absence of PA and positive IFA in Helicobacter pylori-associated pangastritis. 8This is due to a reduction in gastric acid secretion and subsequent reduced release of cobalamin from haptocorrin to the intrinsic factor. 8ence, true cases of H. pylori-associated pangastritis may be missed in the presence of IFA. 9 Although the issue of gastritis is not addressed in the Manchester Anaemia Guide, patients with vitamin B12 deficiency, with or without associated anaemia, should undergo gastroscopy with bioptic sampling of antral, corporal, and duodenal mucosa at least once during the diagnostic work-up to definitely exclude the presence of atrophic gastritis, H. pylori-associated pangastritis (which is highly reversible with H. pylori eradication treatment), or coeliac disease. 102] Oral cyanocobalamin has been proposed to be a cheaper and equally effective option to intramuscular hydroxocobalamin. 2,13owever, parenteral therapy remains the most common 14 and this was observed in our audit population.Nyholm et al. 2 concluded that oral cyanocobalamin was an effective alternative to injections for B12 deficiency of most causes, but this is rarely used in the United Kingdom and is only available to vegans or patients with dietary B12 deficiency on an NHS prescription.Oral B12 also provides patients with an alternative choice, and helps reduce costs in primary care and the burden of work for nurses either at home or at the surgery in giving B12 injections.Moreover, it has been proposed that oral B12 can be used to treat B12 deficiency related to H. pylori infection. 9However, patients with a rare condition known as tobacco amblyopia should remain on hydroxocobalamin injections as oral cyanocobalamin may worsen the condition. 15e option of oral tablets or other routes of therapy could be offered to patients, but determining the optimal dosage to administer may be difficult, especially for older patients who may need oral dosages of more than 500 μg/day for optimal absorption. 16t this point, screening is controversial and the evidence for it is scarce.There have been debates over the value of testing where there is an apparent aetiology (e.g.gastric surgery), as treatment is similar irrespective of the cause. 7However, clinicians should identify patients at risk of developing B12 deficiency and consider testing for it.

Recommendations
Table 2 summarizes the recommendations for change.We aim to re-audit this study in 2 years.• Some patients are put on long-term B12 injections inappropriately.
• Some patients are put on a different type of B12 therapy (i.e.intramuscular instead of oral).
• Review the patients who are not managed according to guidelines -change regime as appropriate and review their B12 prescriptions.
• Use oral supplements in patients who are B12 deficient from a dietary cause.
3. The causes of some patients' vitamin B12 deficiency are unknown.
• Review these patients, chase up results, and perform the necessary investigations (e.g.IFA test, anti-parietal cell antibodies test).
• In all future diagnostic work-ups for suspected B12 deficiency, PA should first be excluded.4. Patients did not have a follow-up monitoring of serum B12 levels.
• Check serum B12 levels for patients on long-term metformin and non-vegan patients with dietary B12 deficiency.Document this and arrange for 6monthly monitoring.
• Review these patients and cease B12 therapy if necessary.

Documentation issues:
• Cause of B12 deficiency unclear.
• Dietary status of patients unclear.
• Future documentation to include: a. Cause of B12 deficiency.b.Appropriate B12 treatment regime plan.c.For patients with a dietary B12 deficiency: i. Dietary status (e.g.vegan or nonvegan), ii.
Any dietary advice given.
• Re-document all current patients with vitamin B12 deficiency.

Conclusion
Clearly, there needs to be improvements in performing IFA tests for patients with established vitamin B12 deficiency, as well as in documentation, assessing adherence to management and ensuring the implementation of monitoring guidelines.With these recommendations in place, we foresee that more patients with PA will be identified, thus rectifying any diagnostic doubts and allowing for the evaluation of the need for specialist referral or changes in treatment.These measures may result in safe prescriptions and potential cost savings, and improved documentation will allow clinicians to be aware of the patient's diagnosis and treatment plan and this, in turn, will improve follow-up monitoring.

Figure 1a . 3 Figure 1b .Figure 2 .Figure 3A .
Figure 1a.Management of vitamin B12 deficiency in patients with/without neurological symptoms.Adapted from ref. 3 No. of patients on current vitamin B 12 replacement (identified on Emis) = 38 No. of patients with dietary only B 12 deficiency = 6 (16%) No. of patients with non-dietary B 12 deficiency

Table 1 .
Summary of audit results