Article Text

Computer-based-limited and personalised education management maximise appropriateness of vitamin D, vitamin B12 and folate retesting
1. Michela Pelloso1,
2. Daniela Basso1,
4. Paola Fogar1,
5. Mario Plebani1,2
1. Correspondence to Dr Michela Pelloso, Department of Laboratory Medicine, University-Hospital of Padova, Via Giustiniani 2, Padova 35128, Italy; michela.pelloso{at}gmail.com

## Abstract

Aim To identify the best management strategy for improving the appropriateness of vitamin D, vitamin B12 and folate retesting.

Methods The study was conducted between 3 November 2012 and 8 June 2015, with inpatients and outpatients being considered separately. After an observational reference period (3 November 2012 to 14 September 2013), an information technology (IT)-based permissive strategy (16 September 2013 to 27 July 2014) followed by a limiting strategy was used to manage the demand for inpatient retesting. For outpatients, an educational strategy period (28 July 2014 to 16 December 2014) with direct contact between medical personnel and general practitioners (GPs) was followed by a post-educational period without any restriction. Data from a total of 66 496 patients for vitamin D, 14 618 for vitamin B12 and 14 445 for folate were retrieved from the laboratory IT system. The main outcomes measures were inappropriate vitamin D, vitamin B12 and folate retesting. The minimal retesting intervals were 90 (vitamin D) or 180 days (vitamin B12 and folate).

Results In the absence of a laboratory demand strategy, the frequency of inappropriate retesting for vitamin D, vitamin B12 and folate was 60%, 94% and 93%, respectively, for inpatients, and 27%, 87% and 87%, respectively, for outpatients. A limiting IT-based demand management strategy reduced inappropriate retesting for vitamin D (36%), but not for vitamin B12 and folate. The educational strategy was followed by a reduction in inappropriate retesting among outpatients (16% for vitamin D, 72% for vitamin B12 and folate).

Conclusions Laboratory demand management based on an IT-limiting management strategy or on education of the referring physicians appears helpful in maximising appropriate retesting.

• LABORATORY MANAGEMENT
• VITAMIN D
• COMPUTER ASSISTED
• EDUCATION
• LABORATORY TESTS

## Introduction

Reduction in worldwide healthcare budgets has led to pressure to reduce healthcare costs. Although clinical laboratories account for no more than 4% of total healthcare expenses, in industrialised countries they are the main target for cost containment.1 ,2 This is probably owing to the enormous number of available clinical laboratory tests, their clear cost for healthcare and the immediate effect of any cost cutting.3 This might explain the clinical decisions made by administrators, who might forget such clinical laboratory tests are indispensable for clinical decision-making.1 ,4 This fact is fundamental to any ‘demand management’ intervention designed to maximise the use of appropriate laboratory tests as the prerequisite for improving the management of patients, thus allowing cost saving.5 Causes of inappropriate requesting include wrong patient (eg, digoxin determination in patients not taking the drug), wrong test (eg, coagulation factor VII in patients receiving warfarin), wrong time (eg, circadian rhythm hormones outside the peak or trough levels) and wrong process (eg, determination of cytokines in non-refrigerated samples).3

Inappropriate testing leads to test overuse and underuse, both of which compromise the quality of healthcare.6 The former can be managed, but the latter is more difficult to identify and improve, since it calls for knowledge of patients’ clinical situations and needs.7 Inappropriate retesting, a problem in test overuse, may occur owing to (1) lack or ignorance of guidelines, (2) a search for any possible support in cases of diagnostic uncertainty, (3) pressure from patients, (4) ‘defensive’ medical practice and/or (5) negligence in previous data collection and analysis.3–4 ,8 In 2013, in order to provide assistance in managing retesting, the Association for Clinical Biochemistry and Laboratory Medicine and the Royal College of Pathologists released the national minimum retesting interval (NMRI) project, which specified how often, if at all, a test should be repeated, based on physiological properties, biological half-life, analytical aspects, treatment and monitoring requirements and established guidelines.9

Management of retesting demand can have different approaches, ranging from weak strategies, such as education initiatives or distribution of guidelines, to strong strategies which are often based on information technology (IT) tools, including laboratory formulary, order entry design, algorithms and reflex testing protocols, clinical pathology consultancy services and finance-based restrictions.2 ,3 ,6 ,11–13

An area for demand management intervention in testing relates to vitamin D, vitamin B12 and folate, which are among the nine vitamins crucial for human health.14 Demand for vitamin D testing, in particular, has grown exponentially in the last decade,15 and its overuse is probably related to the initial testing and also to inappropriate retesting, which in turn causes an unnecessarily increased use of dietary products supplemented with vitamin D. The cost for these has risen in the USA from approximately $108 million in 2007 to$713 million in 2013.16

Deficiency in vitamin D, long known to be essential for healthy bones,17 is now considered a risk factor for several diseases, including neurodegenerative disorders and cancer.18–22 In mammals, vitamin D is primarily produced in the skin upon exposure to UV light; diet is another source of the vitamin. The serum levels of vitamin D reflect the concentration of 25-hydroxyvitamin D, the most stable circulating form of the vitamin, which depends on the total supply of vitamin D.16 Although universal screening for vitamin D status is not recommended, it is indicated in certain clinical settings, especially for frail (eg, elderly) patients. When vitamin D deficiency (<30 nmol/L) is found, supplementation is recommended in an age-related dosage.23 However, when baseline vitamin D concentrations are low, levels of 75–100 nmol/L are unlikely to be achieved in the short term, and even with high dosage supplementation (800 IU/day), vitamin D serum levels reach a plateau after about 60 days’ administration.24 Based on this pharmacodynamic knowledge, the final consensus recommendations of the NMRI project are that vitamin D retesting should be performed no sooner than 3 months after starting vitamin supplementation.9

The biochemical pathways of vitamin B12 and folate are closely intertwined. Therefore, these vitamins are usually assayed in serum together, as patients with either type of deficiency present similar clinical features, mainly megaloblastic anaemia and neuropathy.25 The British Committee for Standards in Haematology recommends repeat serum vitamin B12 retesting no sooner than 1–2 months after the initial testing to confirm the presence of reduced values (110–148 pmol/L or 150–200 ng/L),25 while the final consensus recommendations of the NMRI project are that retesting in cases of vitamin B12 and folate deficiency is unnecessary.9 In cases of cobalamin therapy, retesting is recommended no earlier than 3–4 months after the end of 4 weeks’ low-dose oral treatment.25 Oral supplementation is recommended in the presence of folate deficiency (<7 nmol/L or 3 µg/L). Since a decline in serum folate levels appears to occur several weeks after deprivation, increased serum levels after supplementation are also likely to take several weeks before becoming detectable.25

The above observations indicate that it is inadvisable to perform vitamin D, vitamin B12 and folate retesting outside the recommended intervals. Since the referring physician for outpatients is almost always the same person, we felt that a strong education-based strategy might improve the appropriateness of retest requests from general practitioners (GPs) and ambulatory care physicians. In hospitals, on the other hand, patients’ stay usually does not exceed the minimum retesting interval (MRI) for vitamins and several hospital physicians might be involved in treatment; we therefore hypothesised that IT-based strategies in demand management might be more effective. In this study we compared IT-based weak (pop-up alert) with strong (ban) demand management strategies for hospitalised patients, whereas for outpatients we assessed whether a strong educational strategy would lead to appropriate retesting of vitamin D, vitamin B12 and folate in a large-scale clinical laboratory.

## Methods

### Setting

This study was conducted at the Department of Laboratory Medicine of the University-Hospital of Padova. This department, which operates 24 h a day 7 days a week (365 days a year) performs more than 34 000 tests a day, serving both inpatients and outpatients. The University-Hospital of Padova is a large academic medical institution serving as a tertiary care centre of more than 500 000 people from Padova city and the surrounding area, in northeast Italy. This healthcare institution with 1400 beds provides specialty care in surgical, medical, oncological and maternal-child areas, with 298 emergency room admissions, 189 ordinary hospitalisations and 1346 specialist examinations every day. Outpatients, supported by a network of 200 GPs, comprise 700–1000 people referring to 20 phlebotomy centres of the Department of Laboratory Medicine every day.

### Patients

Inpatients and outpatients were considered as separate groups, the former being peer grouped and considered as pertaining to the medical, surgical, oncological and maternal-child areas in accordance with the University-Hospital of Padova's classification of care units. The outpatient group comprised patients referred by their GPs and those from the ambulatory care clinic.

### Data collection and study design

Vitamin D (chemiluminescent immunoassay, Liaison(R) 25-hydroxyvitamin D total assay, DiaSorin, Saluggia, Italy), vitamin B12 and folate (chemiluminescent immunoassays, Cobas(R), Roche Diagnostics, Germany) data were drawn from the laboratory IT system (TD-Synergy, Technidata Medical Software, Montbonnot, France). The following data were collected: total number of initial tests, and total number of repeat tests, time interval between repeat requests and, for each patient, the baseline and repeat values, age, gender and source (hospital ward for inpatients and ambulatory care clinic or GPs for outpatients). A MRI of 90 days for vitamin D and 180 days for vitamin B12 and folate were defined. When repeat tests were performed before the respective MRI, they were classified as inappropriate.

The flow-chart diagram of the study is shown in figure 1. For reference purposes, we considered data from 3 November 2012 to 14 September 2013, when test requests were not submitted for laboratory management (reference period abbreviated as Ref).

Figure 1

Flow-chart diagram of the study.

Inpatients demand management: from 16 September 2013 to 27 July 2014, an automated IT-based MRI alert (pop-up) was used for inpatients to warn clinicians of inappropriateness of requested retesting (permissive strategy). During the permissive strategy, physicians might stop or complete the inappropriate test request despite the warning pop-up. In the latter case an automatic comment appeared in the results report so that the doctors received information enabling them to make more appropriate requests in the future. Online supplementary figure S1 shows the flow-chart diagram of IT-based pop-up alerts and reports the accompanying warning comments. From 28 July 2014 to 08 June 2015, a limiting strategy for inpatients was applied by introducing an IT-based block to regulate retesting. Only after direct contact between requesting clinicians and laboratory medical staff could the block be overcome.

Outpatients demand management: an educational strategy was applied from 28 July 2014 to 16 December 2014. If inappropriate retesting was requested, the analysis was IT-blocked and replaced with a personalised comment manually introduced by laboratory medical personnel, who also advised GPs by phone (educational strategy). In both the personalised comment and the phone call, the GP was informed of previous results, and the correct time interval for performing appropriate retesting. To verify whether the educational strategy was effective, from 17 December 2014 to 08 June 2015 requests for outpatient retesting were not submitted for laboratory management (post-educational strategy).

### Interviews of clinicians

On completing the limiting strategy for inpatients and the post-educational strategy for outpatients, personnel requesting retesting (50 internal physicians and 50 GPs) were interviewed to ascertain whether the rate of inappropriate testing was correlated with their beliefs or practices (details of questions are given in the online supplementary methods).

### Statistical analysis

Statistical analysis of data was carried out using the χ2 test, one-way analysis of variance and Bonferroni's test for pairwise comparisons, the Kruskal–Wallis test, the Wilcoxon signed rank test and binary logistic regression analysis (StataSE 13, StataCorp Texas, USA).

## Results

### Inappropriate retesting is frequent, occurring mainly among inpatients

Table 1 shows the total number and percentages of inpatients and outpatients for whom vitamin D, vitamin B12 and folate were tested only once or at least twice during the 10-month reference observation period.

Table 1

Baseline values in the reference period (3 November 2012 to 14 September 2013)

We then focused on patients who underwent repeated testing by analysing whether during the reference observation period vitamin D, vitamin B12 and folate retesting was appropriate or inappropriate. Retesting was considered inappropriate when the time interval between the first assay and retesting was ≤90 days for vitamin D and ≤180 days for vitamin B12 and folate. Inpatients and outpatients were considered separately and, among inpatients, the hospital area from which they came was also taken into account (table 2).

Table 2

Inappropriate vitamin D, vitamin B12 and folate retesting among inpatients and outpatients in the reference period (3 November 2012 to 14 September 2013)

On considering the reference period, we ascertained whether there was any significant difference between the first and the second value for vitamin D, vitamin B12 and folate when retesting was appropriate or inappropriate among inpatients or outpatients (see online supplementary table S1).

Binary logistic regression analysis was performed considering vitamin D, vitamin B12 and folate retesting in order to ascertain whether age, gender, first test result and inpatient or outpatient status led to inappropriate retesting in the reference period (table 3).

Table 3

Binary logistic regression analysis. Appropriate versus inappropriate vitamin D, vitamin B12 and folate retesting were dependent variables, whereas age, gender, inpatient/outpatient status and first test results were predictors

### Restrictive management and education lead to appropriate retesting

The effects of permissive and limiting strategies in reducing inappropriate retesting were analysed by binary logistic regression analyses, made on inpatients (table 4).

Table 4

Effects of permissive and limiting strategies in reducing inappropriate retesting among inpatients (binary logistic regression analyses)

The efficacy of the educational strategy in reducing inappropriate retesting was analysed by binary logistic regression analyses, made on outpatients (see online supplementary table S2).

Figure 2 shows the percentage of appropriate retesting (columns) and the total number of patients who underwent vitamin D, vitamin B12 and folate testing once (squares with connecting lines) or at least twice (dots with connecting lines) when the different strategies for demand management were applied.

Figure 2

Demand management strategy and appropriate retesting. Vitamin D (A and B), vitamin B12 (C and D) and folate (E and F) data are shown for inpatients (A, C, E) and outpatients (B, D, F).The percentage of appropriate retesting (bars) and the total numbers of patients who performed the test only once (no retesting, squares with connecting lines) or at least twice (retesting, dots with connecting lines) are shown. Educational, personalised comment manually introduced by laboratory medical personnel, who also advised general practitioners by telephone; Limiting, IT-based limiting strategy; Permissive, IT-based permissive strategy; Post-Educational, absence of any demand management strategy; Ref, reference period.

### Clinicians’ beliefs expressed in interviews

Almost 40% of the hospital physicians declared that they were aware of vitamins’ MRI before laboratory demand management. Only 10% of physicians considered the permissive strategy conducive to learning. The reasons for inappropriate retesting despite previous or acquired knowledge of MRI were mainly (1) the non-immediate availability of previous patient data and (2) the transmission of test requests from physicians to nurses who transferred them into the IT system. Almost all hospital physicians believed that an IT-based block (limiting strategy) was the best available demand management strategy to reduce inappropriate retesting of vitamins.

Data from the questionnaire showed that 81% GPs were unaware of vitamins’ MRI before laboratory intervention and this percentage decreased to 50% after the educational period. In the post-educational period, only 7% of GPs declared that they made appropriate requests for vitamin retesting (odds between post-educational and ref=0.320) and the reasons for this inappropriateness were wasteful organisational set-up of GP’s office (32%), pressure from patients that was often driven by specialists’ requests (58%) and other unspecified reasons (10%). Almost 90% of GPs believed that a laboratory IT-based intervention was a useful initiative, but only when implemented in association with a personalised comment reporting information on previous tests performed (date and value) and appropriate MRI.

## Discussion

Industrialised countries must now deal with the great challenge of providing high-quality care with reduced resources. Medical laboratories are often in the front line for cost reduction policies because they are an easy target for restrictions and limitations.1 ,2 Indeed, with their enormous throughput, medical laboratories might be likened, by an inexpert eye, to a mass production industry. Yet around 70% of all medical decisions are based on laboratory test results,1 the overall use and interpretation of which is guided by shared protocols/guidelines, and often several other factors, including the rapid expansion of diagnosis, monitoring, screening and prognosis testing, changes in clinical practices, patient empowerment, an ageing population, a growing number of patients with chronic diseases, defensive testing and, last but not least, ignorance of diagnostic significance, sensitivity and specificity. All these factors might lead to inappropriate testing, while appropriate testing is considered crucial to reducing costs and improving quality in laboratory medicine demand management.5

The aim of this study was to identify a strategy to manage the growing demand for vitamin retesting. The relevance of vitamin D, vitamin B12 and folate retesting emerged clearly from data obtained in the reference period: in a 10-month period, more than 10% of patients underwent repeat vitamin B12 and folate testing and more than 15%, repeat vitamin D testing. These percentages, although higher than the threshold of 7.4% defined as a cut-off point for overuse,8 do not indicate whether the retestings were appropriate or inappropriate. To investigate this, retesting was classified as appropriate or inappropriate based on the test-specific MRI.9 ,25 The rate of inappropriate retesting for vitamin B12 and folate was high among both outpatients (87%) and inpatients (>93%), and for vitamin D mainly among inpatients (60%), but also among outpatients (27%). The high overall rate of inappropriate vitamin retesting among inpatients was common to the different hospital areas and to outpatients in general practice and ambulatory care clinics.

We then investigated factors that might be responsible for inappropriate retesting, considering age, gender, first test result and hospitalisation, patients’ data being retrievable from our IT system. In our analysis, the lack of data on treatment, comorbidities and other potential clinical decision-making information was a limitation, but it was partly overcome by the enormous number of observations made. Inappropriate vitamin B12 and folate retesting depended significantly on inpatient status, whereas age, gender and first test result were not significant. Inappropriate vitamin D retesting depended on a more complex interplay among predictor variables. Appropriate vitamin D retesting correlated with outpatient status (4.68-fold higher than for inpatients), first test result (1% increase for any nmol/L vitamin D increase), decreasing age (1%) and female gender (18%) (table 3). Accordingly, lower median values for the first vitamin D test result were found among inpatients and outpatients who underwent inappropriate retesting than among those who underwent appropriate retesting, although this difference was not seen  for vitamin B12 and folate (see online supplementary table S1). If the decision to repeat vitamin D testing earlier than recommended is partly determined by the first test result, one might expect significant variations between the first and second values among patients with inappropriate rather than appropriate retesting. However, this was not confirmed, suggesting that the choice of retesting time for vitamin D depends on numerous factors, indicating the need for effective management.

Retesting management may follow more or less stringent lines.2 ,3 ,6 ,10–13 We decided to start demand management with less stringent criteria, although the management strategies for inpatients and outpatients were different. For inpatients computerised order entry-based interventions were adopted, while for outpatients direct contact between laboratory staff and GPs was the basis for education-based interventions. For all vitamins we showed that a moderate-strength strategy reduced inappropriate retesting but its efficacy was limited: only a 7% increase in appropriate retesting for vitamin D and vitamin B12, and a 5% increase for folate. This indicates that ‘pop-up window fatigue’ does not enhance appropriateness, as observed by Procop et al26 ,27 in their study on duplicate testing. A more stringent computer-based strategy effectively increased the percentage of appropriate vitamin D retesting among inpatients (64% in the limiting period with respect to 40% reference), and limited the increasing trend in the total number of patients for retesting (see figure 2A), which parallels the overall increase in vitamin D orders.15 Similarly, the limiting strategy caused a drop in the number of vitamin B12 and folate retests (from about 200 in the permissive to 55 and 87, respectively, in the limiting period), but the percentages of inappropriate retesting for these two analytes remained unchanged or increased, thus explaining the lack of significance with the logistic regression analysis. This finding, together with the lack of 100% appropriateness in vitamin D retesting in the limiting period, occurred because restriction was not absolute, being overruled when laboratory staff recognised the clinical need, which occurred mainly in the oncological and maternal-child hospital areas to which frailer patients are referred. The educational strategy involving direct contact between the laboratory staff and general practitioners improved appropriate retesting for all vitamins, and was found to significantly ameliorate physicians’ practice. In the post-educational period a fall in the number of retests was seen also among outpatients despite the absence of any computer-based or laboratory staff-based interventions. However, in the post-educational period, inappropriate retesting was unexpectedly light and, according to GPs due to practical reasons, mainly waste cabinet organisation and pressure from patients. Interestingly, at the end of the study, both hospital physicians and GPs preferred IT-based limiting strategies for reducing inappropriate retesting.

Both hospital physicians and GPs interviewed agreed that IT-based strategies would be much more effective if the pop-up warning also reported the most recent result. The absence of this was one limitation of our study. These indications allow transferability of the results of this study to other countries and also to other clinical scenarios, such as medium- or low-volume healthcare institutions.

Demand management should, by improving appropriateness, reduce overall costs. For inpatients, the total numbers for vitamin D, vitamin B12 and folate testing, including retesting, increased from 8317 to 9314 from the reference to the permissive periods with an overall trend of 12% for increased test demand. This can be translated into an increase of about 14% (€13 000; June 2015: 1 €=US\$1.12), calculated on the basis of reimbursement tables. Assuming a constant trend, the expected total costs for these tests would be in the region of €120 000 during the limiting period, while application of demand management would incur a total cost of €94 370 with a 21% saving in costs.

In conclusion, inappropriate retesting of vitamins, an open and relevant area in laboratory medicine, could be effectively managed by both computer-based and educational interventions. The success of this approach will depend mainly on the underlying collaboration between laboratory staff, the hospital itself and GPs.

### Take home messages

• Vitamin D, vitamin B12 and folate testing is relevant in clinical care, but inappropriate retesting might have a negative effect on patient management in view of increasing healthcare costs.

• Information technology-based minimal retesting interval strategies effectively increase the percentage of appropriate retesting and limit the increasing trend for retesting of inpatients.

• An educational strategy, involving direct contact between the laboratory staff and general practitioners, improves appropriate retesting for all vitamins in outpatients.

• Appropriate testing is considered a key to reducing costs and improving quality in laboratory medicine, and demand management application is crucial to achieving this goal.

## Acknowledgments

The authors thank Mr Angelo Legnaro, Mr Alessio Pinato and Mr Gianni Vecchiato for their support in the development of computer-based rules. The authors also thank Sara Jane Pearcey for her valuable help with the English.

• ## Abstract in Italian

This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

• ## Supplementary Data

This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

## Footnotes

• Handling editor Tahir Pillay

• Contributors Conceived and designed the study: MPe, DB, MPl. Collected and analysed the data: MPe, DB, AP, PF. Wrote and edited the paper: MPe, DB, AP. Revised the work: PF, MPe. All authors read and approved the final manuscript.

• Competing interests None declared.

• Provenance and peer review Not commissioned; externally peer reviewed.

## Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.