Potential over request in anemia laboratory tests in primary care in Spain

Abstract

Objectives

The aim was to study the inter-practice variability in anemia laboratory tests requested by general practitioners in Spain, to evaluate for a potential requesting inappropriateness.

Methods

Laboratories from diverse Spanish regions filled out the number of cell blood count, ferritin, folate, iron, transferrin, and vitamin B12 requested by general practitioners during 2012. The number of test requests per 1000 inhabitants and ratios of related tests requests were calculated. The results obtained in hospitals from different areas (urban, rural, or urban–rural), type of management (public or private), and geographic regions were compared.

Results

There was a high variability in the number of test requests and ratios of related tests. Cell blood count was over requested in rural areas and in hospitals with private management. Andalucía was the community with the lowest number of iron requests and the lowest folate/vitamin B12 indicator value.

Conclusions

Iron and transferrin seemed over requested in some areas; as were folate and ferritin when compared to vitamin B12 and cell blood count, respectively. The differences observed between areas indicate that other factors besides clinical reasons could be behind that variability and emphasize the need to accomplish interventions to improve the appropriate use of anemia laboratory tests.

Keywords:

• Anemia
• Appropriate use
• General practitioners
• Indicators
• Laboratory professionals
• Laboratory tests utilization

Introduction

Anemia is the most common hematological disorder in the elderly, it is a marker of mortality and morbidity, it adversely affects quality of life, and can be a symptom of a severe underlying disease such as malignancy.¹ It is a condition associated with a worse prognosis and especially higher mortality, in various disorders. In all, anemia is a common problem in primary care, and in general, its management is a main challenge to the health-care system.²

It is the proper attitude of general practitioners (GPs) and laboratory professionals that will usually lead to its diagnosis. The diagnosis of anemia is indeed based on the laboratory; it is defined as a hemoglobin (Hb) concentration below a lower reference value. To make a generalized approach, the World Health Organization established a reference range for normal blood Hb concentration, depending on age and sex.³ Consequently, a simple laboratory algorithm on the Hb result may be applied as a guide to diagnose the disorder.

The red cell volume (RCV) is a central value to classify and define the type of anemia. Among the low RCV value situations, iron deficiency anemia (IDA) is the most common form worldwide.⁴ Traditionally, measurements of iron and iron-binding capacity were performed by laboratories for the diagnosis of IDA. However, today the optimal diagnostic approach is to measure serum ferritin as an index of iron storage.⁵ Ferritin provides the best means to confirm IDA and is superior to iron, iron-binding capacity, or transferrin saturation for this purpose.^6,7 In fact, ferritin is now the most reliable test for iron deficiency, although its values may be influenced by the presence of acute or chronic inflammation. In addition, in the majority of cases, serum ferritin provides a simple method to discriminate between IDA and anemia of chronic inflammation⁶ and advises the GPs to perform additional laboratory tests in patients with decreased Hb, to establish the cause of the anemia.

In the setting of low Hb values, if RCV is increased, vitamin B12 and/or folate deficiency will be the most common cause of macrocytic anemia in western populations.⁸

In all, an appropriate management of anemia laboratory tests is crucial for a correct diagnosis and management of the disease. Nowadays, through the study of geographical differences,^9–11 we can investigate the variability in tests requesting and deduce the root causes of a potential inappropriate demand.

The aim of this research was, first, to study the inter-practice variability in anemia laboratory tests requested by GPs in Spain; second, to compare and analyze indicators results according to hospital characteristics; and finally, to try to find out the degree in requesting inappropriateness.

Material and methods

Data collection

Encouraged by the previous pilot studies in the Valencian community⁹ and all around Spain,^10,11 a call for data was posted via email. One hundred and forty-one Spanish laboratories were invited to participate, to fill out an enrollment form and submit their results online. We obtained production statistics (number of tests requested by GPs) for the year 2012 from laboratories from diverse regions across Spain. Every patient seen in any primary-care center, regardless of the reason for consultation, gender, or age, was included in the study. Each participating laboratory was required to be able to obtain patient data from local databases and to provide organizational data (i.e. population served, public/private management, location). Cell blood count (CBC) panel and serum ferritin, folate, iron, transferrin, and vitamin B12 tests requests were examined in a cross-sectional study.

Data processing

After collecting data, two types of appropriateness indicators were calculated: the number of each investigated test (CBC, ferritin, folate, iron, transferrin, vitamin B12) per 1000 inhabitants; and ratios of related tests requests (ferritin/CBC, transferrin/CBC, and folate/vitamin B12, and the index of variability) calculated as top decile divided by bottom decile (90th percentile/10th percentile).

The indicator results obtained in the laboratories in the three regions with the higher number of areas participating in the study were compared between them and to the pooled results of the remaining regions in order to establish whether there were regional differences in the requesting patterns. Results were compared with the median age of the population, gross domestic product (GDP) per inhabitant, and the unemployment rate of each region.¹²

Finally, we calculated whether the rate of test requests was different according to the setting (rural, urban, or rural–urban) and whether the institution had a public or private management.

Statistical methods

The statistical treatment of the calculated data included: the distribution, the mean, 95% confidence level for the mean, standard deviation, median, and interquartile range. The analysis of the distribution of the number of test requests per 1000 inhabitants was conducted by way of the Kolmogorov–Smirnov test.

The differences in the indicators results according to the hospital characteristics and per region were calculated by way of a Kruskal–Wallis test analysis.

A two-sided P ≤ 0.05 rule was utilized as the criterion for rejecting the null hypothesis of no difference.

Results

We obtained production statistics from 64 laboratories at different hospitals from diverse regions across Spain. A total of 14 846 065 patients, from 12 Spanish Communities were included (Valencian community, 22 laboratories and 4 703 737 patients; Andalucía, 10 laboratories and 2 791 052 patients; Castilla y Leon, 8 laboratories and 1 601 394 patients, and remaining regions 24 laboratories and 5 749 882 patients). Table 1 displays a summary of the organizational data of the different laboratories that participated in the study.

Table 1. Descriptive characteristics of the hospitals/health-care departments that participated in the study

The number of different tests ordered is shown in Table 2. CBC, ferritin, iron, folate, and transferrin were 3, 6, 88, 75, and 260 fold over requested in certain areas. Ratios of related tests request also showed a high variability. In total, approximately one ferritin test was demanded for every three CBCs. The folate/vitamin B12 indicator result showed approximately a one-to-one requesting ratio.

Table 2. Descriptive statistical analysis and variability index (percentile 90/percentile 10) of every indicator result

No significant differences were noted in test requesting in rural, urban, or rural–urban locations except for CBC per 1000 inhabitants that was higher in rural than in rural–urban areas (Table 3). The indicator was also higher in centers with private management, when compared to public (Table 3). The regional differences in the indicator results are shown in Table 4. The region with the lowest GDP and highest unemployment rate had the lowest request rate of all anemia laboratory tests (Table 5).

Table 3. Differences of appropriateness indicators results obtained at the laboratories of the different locations and management

Table 4. Differences of appropriateness indicators results obtained at the laboratories of the different regions of Spain

Table 5. Indicator of region in 2012

Discussion

There were significant differences in the request of laboratory tests regarding anemia, even in common tests such as CBC which was three times more requested in certain areas as compared to others. No differences were observed in request behavior between type of location and management, except for CBC per 1000 inhabitants. The observed high variability and the current evidence in the appropriate utilization of iron and transferrin in anemia show that both tests may be over requested in some areas. A request for approximately one ferritin for every three CBCs showed that ferritin could also be over requested.

The higher CBC request in rural areas and hospitals with private management could be explained by a lower patient referral to the hematologists or by the fact that GPs deal with more specialized pathologies in such regions. However, more studies would be needed to find out the real explanation of these observed differences. The higher CBC requesting in Valencian community could be explained by the fact that it was in this the Spanish community where hospitals with private management were located.

After CBC, ferritin was the highest requested test, as IDA is the most common form of anemia.⁴ However, the observed request seems too high; in fact, more than one ferritin for every three CBC was ordered. Besides, a redundant test, transferrin, was also highly requested and with a high variability. The use of transferrin saturation in the diagnosis of IDA has been discouraged.^13,14 Currently, ferritin is considered as the first marker for IDA, and iron and transferrin are considered as redundant tests.^6,7 This fact and our results suggest an inappropriate transferrin over requesting in some areas. Iron could also be over requested in some areas. Nowadays, the indication for the analysis of serum iron saturation is limited to hemochromatosis.¹⁵

It is important to reach an agreement in the selection of tests to diagnose fairly well-established clinical entities such as anemia.¹⁶ In fact, if over requested, some tests results are very seldom outside the reference range (e.g. vitamin B12 and folate) unlike other tests such as serum iron that have a high intra-individual variability.¹⁷

Folate/vitamin B12 ratio was around 1 in all laboratories, despite the additional indications of vitamin B12 in primary-care setting, such as neuropsychiatric symptoms in older people.⁸ Moreover, serum folate determination is of little clinical significance in patients without known risk factors for folate deficiency.¹⁸ Our study results could suggest a folate inappropriate over request.

The differences observed between regions in almost every indicator studied show that additional reasons different to case mix patients could influence the GPs tests request behavior, and that interventions in collaboration with every stockholder that intervene in anemia patients management are necessary to improve disease management.

The study had certain limitations. The differences in anemia tests laboratory request between health-care regions in Spain could be partly explained by case mix variations in the different areas, the possible variability in the patient population or by regional different degrees in the management of anemia disorders in primary care.

The study results highlight the large variability in anemia laboratory tests requested in primary care in Spain. The appropriateness indicator results suggest an iron, transferrin, and folate over request. The differences observed between the different regions suggest that other factors independent of clinical reasons could have been behind that inappropriateness.

Disclaimer statements

Contributors MS, ML-G, EF, JU and CL-S conceived, designed and drafted the manuscript; MS, ML-G, EF, JU and CL-S revised the final manuscript. All authors contributed to the concept, reviewed all versions of the manuscript and commented, and approved the final revised version of the manuscript.

Funding This study was supported by research funding: Ignacio H. de Larramendi aid to research from Fundación Mapfre.

Conflicts of interest None.

Ethics approval This study was approved from local Research Ethics Committee.

Acknowledgements

Members of the REDCONLAB working group are the following (in alphabetical order): Alfonso Pérez-Martínez (Hospital Morales Meseguer); Amparo Miralles (Hospital de Sagunto); Ana Santo-Quiles (Hospital Virgen de la Salud, Elda); Angeles Giménez-Marín (Hospital de Antequera); Antonio Buño-Soto (Hospital La Paz, Madrid); Antonio Gómez del Campo (Complejo Asistencial de Ávila); Antonio León-Juste (Hospital Juan Ramón Jiménez, Huelva); Antonio Moro-Ortiz (Hospital de Valme, Sevilla); Begoña Laiz (Hospital Universitario y Politécnico La Fe de Valencia); Berta González-Ponce (Hospital Da Costa, Burela); Carmen Hernando de Larramendi (Hospital Severo Ochoa de Leganés); Carmen Vinuesa (Hospital de Vinaros); Cesáreo García-García (Hospital Universitario de Salamanca); Consuelo Tormo (Hospital General Universitario de Elche); Cristina Santos-Rubio (Hospital Río Tinto, Huelva); Cristóbal Avivar (Hospital de Poniente, El Ejido); Diego Benitez Benitez (Hospital de Orihuela); Eduardo Sanchez-Fernandez (Hospital del Vinalopo, Elche); Emilia Moreno-Noguero (Hospital Can Misses); Enrique Rodríguez-Borja (Hospital Clínico Universitario de Valencia); Esther Roldán-Fontana (Hospital La Merced. Area de Gestión Sanitaria Sevilla Este); Fco. Javier Martín Oncina (Hospital Virgen del Puerto de Plasencia, Caceres); Félix Gascón (Hospital Valle de los Pedroches, Pozoblanco); Fidel Velasco Peña (Hospital Virgen de Altagracia, Manzanares); Goitzane Marcaida (Consorcio Hospital General Universitario de Valencia); Inmaculada Domínguez-Pascual (Hospital General Universitario Virgen del Rocio, Sevilla); Isidoro Herrera Contreras (Complejo Hospitalario de Jaén); Jose Luis Barberá (Hospital de Manises); Jose Luis Quilez Fernandez (Hospital Universitario Reina Sofia de Murcia); Jose Luis Ribes-Vallés (Hospital de Manacor); María Fernández García (Hospital Santiago Apostol de Miranda de Ebro); Ricardo Molina Gasset (Hospital Virgen de los Lirios, Alcoy); Jose Antonio Ferrero (Hospital General de Castellón); Jose Vicente Garcia-Lario (Hospital Virgen de las Nieves, Granada); Juan Ignacio Molinos (Hospital Sierrallana, Torrelavega); Juan Molina (Hospital Comarcal de La Marina, Villajoyosa); Julian Diaz (Hospital Francesc de Borja, Gandia); Laura Navarro Casado (Complejo Hospitalario Universitario de Albacete); Leopoldo Martín-Martín (Hospital General de La Palma); Lola Máiz Suárez (Hospital Universitario Lucus Augusti, HULA, Lugo); M Dolores Calvo (Hospital Clinico de Valladolid); M. Amalia Andrade-Olivie (Hospital Xeral-Cies, CHU Vigo); M. Angeles Rodríguez-Rodriguez (Complejo Asistencial de Palencia); M. Carmen Gallego Ramirez (Hospital Rafael Mendez, Lorca); M. Mercedes Herranz-Puebla (Hospital Universitario de Getafe); M. Victoria Poncela-Garcia (Hospital Universitario de Burgos); M^a José Baz (Hospital de Llerena, Badajoz); M^a José Martínez-Llopis (Hospital de Denia); Mabel Llovet (Hospital Verge de la Cinta, Tortosa); Mamen Lorenzo (Hospital de Puertollano); Marcos Lopez-Hoyos (Hospital Universitario Marques de Valdecilla); Maria Jose Zaro (Hospital Don Benito-Villanueva); Mario Ortuño (Hospital Universitario La Ribera); Marisa Graells (Hospital General Universitario de Alicante); Marta García-Collía (Hospital Ramon y Cajal, Madrid); Martin Yago (Hospital de Requena); Mercedes Muros (Hospital Nuestra Señora de la Candelaria, Tenerife); Nuria Estañ (Hospital Dr. Peset, Valencia); Nuria Fernández-García (Hospital Universitario Rio Hortega, Valladolid); Pilar Garcia-Chico Sepulveda (Hospital General Universitario de Ciudad Real); Ruth Gonzalez Tamayo (Hospital de Torrevieja); Silvia Pesudo (Hospital La Plana); Vicente Granizo-Dominguez (Hospital Universitario de Guadalajara); Vicente Villamandos-Nicás (Hospital Santos Reyes, Aranda del Duero); Vidal Perez -Valero (Hospital Regional de Málaga).