ORIGINALES

 

Assessment of the quality of sample labelling for clinical research

Evaluación de la calidad del etiquetado de muestras para investigación clínica

 

 

Pablo Pérez-Huertas¹, María Tordera-Baviera¹, Concepción Martínez-Nieto², Natalia Benito-Zazo¹, Ana García-Robles¹ and José Luis Poveda-Andrés¹

¹Pharmacy Service, Hospital Universitario y Politécnico La Fe, Valencia.
²Pharmacy Service, Hospital Universitario de la Princesa, Madrid.

Correspondence

 

 

---

ABSTRACT

Objective: To assess the quality of the labels for clinical trial samples through current regulations, and to analyze its potential correlation with the specific characteristics of each sample.
Method: A transversal multicenter study where the clinical trial samples from two third level hospitals were analyzed. The eleven items from Directive 2003/94/EC, as well as the name of the clinical trial and the dose on the label cover, were considered variables for labelling quality. The influence of the characteristics of each sample on labelling quality was also analyzed.
Outcome: The study included 503 samples from 220 clinical trials. The mean quality of labelling, understood as the proportion of items from Appendix 13, was of 91.9%. Out of these, 6.6% did not include the name of the sample in the outer face of the label, while in 9.7% the dose was missing. The samples with clinical trial-type samples presented a higher quality (p < 0.049), blinding reduced their quality (p = 0.017), and identification by kit number or by patient increased it (p < 0.01). The promoter was the variable which introduced the highest variability into the analysis.
Conclusions: The mean quality of labelling is adequate in the majority of clinical trial samples. The lack of essential information in some samples, such as the clinical trial code and the period of validity, is alarming and might be the potential source for dispensing or administration errors.

Key words: Drug labelling; Clinical trials; Medical errors.

---

RESUMEN

Objetivo: Evaluar la calidad de las etiquetas de muestras para ensayos clínicos mediante la normativa vigente y analizar su posible correlación con las características específicas de cada muestra.
Método: Estudio transversal multicéntrico en el que se analizaron las muestras de ensayos clínicos de dos hospitales de tercer nivel. Se estudió la presencia de los once ítems de la Directiva 2003/94/CE, el nombre del ensayo y la dosis en la portada de la etiqueta como variables de calidad del etiquetado. Se analizó la influencia de las características propias de la muestra con la calidad del etiquetado.
Resultado: Se analizaron un total de 503 muestras de 220 ensayos. La calidad media del etiquetado, entendido como el porcentaje de ítems del Anexo 13, fue del 91,9%. El 6,6% no contenía el nombre de la muestra en la cara externa de la etiqueta, mientras que a un 9,7% les faltaba la dosis. Las muestras con presentación de tipo ensayo clínico presentaron mayor calidad (p < 0,049), el enmascaramiento disminuía la calidad (p = 0,017) y la identificación por número de kit o por paciente la aumentaban (p < 0,01). La variable promotor fue la que más variabilidad introdujo en el análisis.
Conclusiones: La calidad media del etiquetado es adecuada en la mayoría de las muestras del ensayo clínico. Resulta preocupante la ausencia de información esencial, como el código del ensayo clínico y el período de validez, en algunas muestras que pueden ser fuente potencial de errores de dispensación o de administración.

Palabras clave: Etiqueta de medicamentos; Ensayo clínico; Errores médicos.

---

 

Contribution to Scientific Literature

This article studies an uncommon but very significant aspect, such as the quality of labelling within clinical trials. There are no articles published in literature to show the quality and quantity of information available in the label, and its potential relationship with dispensing and administration errors. Therefore, this article represents the first publication about the quality of labelling in clinical research.

 

Introduction

Drugs within clinical trials are more complex to dispense and administrate than conventional drugs¹. Besides, the consequences of an error are more dangerous and difficult to detect¹. In this sense, an increasing number of samples require preparation previous to dispensing / administration in the Hospital Pharmacy, which increases the risk of error². The adequate labelling of samples for clinical research is an essential aspect to prevent medication errors, either by patients or by healthcare professionals^3,4,5. Therefore, labels must contain as much information as possible in a clear and concise way, and ensuring an easy search and transmission of information.

The difficulty for patients at the time of looking for information is added up to this higher complexity. Patients won't be able to access their usual means of communication, such as their GP or the community pharmacist. They can only find answers to their doubts if these are collected in the Information Sheet for Patients, or by contacting the Principal Investigator, or the Clinical Trial Unit at the Hospital Pharmacy where they collect their medication. Thus, the label of the clinical trial sample itself becomes their immediate source of information, and sometimes the only one available.

For a medication under clinical research to reach patients, approval must be requested to the regulatory agencies. Among all the information that the pharmaceutical company must send, there is a copy of the label that will be attached to the drug¹. According to Appendix 13 of Directive 2003/94/EC, this label must include up to 11 items (Table 1) in order to meet this regulation. However, the directive also states that any of these items can be omitted, as long as this is adequately justified. This statement is an open door for manufacturers to reduce the information available in the labels, without failing to meet the regulation. An example of justification mentioned by the Directive is the management through a central electronic system of randomization (IVRS or IWRS). The IVRS/IWRS is an interactive voice answer system via telephone or on-line, used to control the telematic management of clinical trials, such as stock management or patient randomization.

 

 

One of the aspects which create a major barrier both for patients and for healthcare professionals is the blinding of the study drugs¹. In order to identify adequately the drug during dispensing, is it essential that the most important information, such as the medication name and dose, appear on the label cover, that is to say, in the most external face of the label. These two significant pieces of data can be omitted to a higher extent in blind samples, therefore increasing the likelihood of dispensing errors.

The primary objective of this study is to assess the quality of labelling for clinical trial samples, understood as the proportion of Appendix 13 items present on the sample label, and the proportion of samples with the number and dose on its outer labelling. Besides, we intended to analyze the impact of the characteristics of each sample on said quality. As secondary objectives, we intended to study the relationship between label quality and management through an IVRS/IWRS system, and the relationship between sample blinding and the presence of the name and dose of the sample on the label cover.

 

Methods

A transversal multicenter study analyzing the quality of labelling for clinical research samples in the Clinical Trial Units of the Hospital Universitario y Politécnico La Fe and the Hospital La Princesa in Madrid. Data were collected since January, 1^st until January, 31^st, 2013, in both hospitals. The inclusion criteria were that the clinical trial they belonged to was still active, and that the sample was present in the unit during the time of the study (Figure 1).

 

 

The variables collected were classified into two groups: sample data and labelling quality data. Sample data included: name of the sample, clinical trial it belonged to, promoter, location at the clinical trial storage, type of presentation (clinical trial sample or relabelled commercial sample), identification (through kit number, randomization number, patient name, or unidentified), blinding, and whether it was managed through a IVRS/IWRS system. In order to determine the quality of labelling, we measured the proportion of items in the Appendix 13 mentioned in Directive 2003/94/EC present at any place in the label (Table 1). Those items split into various different sections were also assessed separately.

On the other hand, the proportion of these items in the primary packaging of the sample was also studied; that is to say, the part of the packaging in direct contact with the medication (blisters, vials, etc.) This part of the study was required because, in some cases, dispensing to patients is conducted after splitting up a package, while preparation in the Pharmacy Unit is always conducted based on this primary packaging.

Besides, data were collected about the number of samples which contained the name of the sample and the dose in the most external face of the label. Given that the majority of relabelled commercial samples are not managed through IVRS/IWRS, it was decided to exclude them from the analysis of secondary objectives, in order to avoid data distortion.

Data collection from samples was conducted by two independent assessment centres. In case of discrepancy, the sample was re-assessed by a third centre. In case of discrepancy in the information about the same sample between both centres, this was also assessed by a third agent. There was an assessment of the quantity of discrepancies for similar samples between both hospitals, in order to measure the degree of subjectivity.

For data collections, a list with the samples recorded in the unit was taken out of the EnsayosPK^® program in the case of La Fe, and from the official program of clinical trials in the case of La Princesa; and a Data Collection Sheet was created. All information was computerized in an Access^® database. Statistical analysis was conducted using logistical regression models, with quality as dependent variable, and sample characteristics as independent variables. This analysis was conducted through version 3.1.2 of the R^® program.

 

Results

In total, 503 samples were analyzed, 429 from the Hospital La Fe and 74 from the Hospital de La Princesa, from 220 clinical trials and 94 different promoters. Table 2 shows the characteristics of the samples studied. No statistically significant differences were found (p=0.378) regarding data collection by the two participating centres for similar samples.

 

 

The mean quality of labelling, understood as the proportion of Appendix 13 items, was 91.9%. Table 3 shows the proportions obtained by sections. Those a) and b) items which were not completely present on the label were assessed separately. This way, 6.2% of samples did not include the promoter's name, 10.9% of samples lacked the promoter's address, and 37.8% lacked their telephone number. Regarding samples with an incomplete item b), 0.2% lacked the pharmaceutical formulation, 0.8% lacked the way of administration, and 0.2% the number of units. From the 503 samples studied, 6.6% did not include the name of the sample on the outer face of the label, while 9.7% lacked the dose. Besides, there were 5.8% of samples which did not include either the name of the sample or its dose.

 

 

Due to the limited number of cases for some variables studied, a variable called Quality Summary was created, represented by the mean proportion of presence of the 11 items in Appendix 13, in order to measure the influence of sample characteristics upon labelling quality. As can be observed in Table 4, the clinical trial-type presentation increases quality to a significant extent (p= 0.049), blinding reduces quality (p = 0.017), and identification by kit number or by patient increases the quality of the label (p< 0.001 and p = 0.003). Besides, the influence of these characteristics of the sample on Appendix 13 items was studied on the samples with label during primary packaging (Table 5). In this case, only the clinical trial-type presentation increased the quality of labelling with statistical significance. On the other hand, the effect of the promoter on label quality was studied, and it was found that this is the factor contributing higher variability to quality.

 

 

 

A 69.6% of those samples with clinical trial presentation were managed through a IVRS/IWRS system. Statistical analysis through logistical regression points out that items b), f), i) and k) are more present when the sample is managed through this system, with statistical significance. On the other hand, it seems that the name and dose are less present on the outer part of the label on IVRS/ IWRS samples, though only with statistical significance for dose (Table 6). On the other hand, samples from blinded clinical trials include a lower proportion of information, name of sample and dose, on the label cover (Table 7).

 

 

 

Discussion

Medication errors have been widely studied during recent years, confirming their importance in number and clinical impact. However, errors associated with drugs in the experimental phase have not been assessed because, being more regulated and individualized, they occur in a lower proportion. Those few studies published reveal that medication errors occur during clinical trials in not inconsiderable proportions⁵. There are even experiences where medication errors have determined the efficacy outcomes of the clinical trial⁷ or errors which have caused deaths among the participants⁸.

In a recent study, a Failure Mode and Effects Analysis (FMEA) was conducted in the area of clinical trials, and it was observed that the processes with the highest risk of error were dispensing and preparation of samples². To prevent errors in the preparation of cytostatic samples for clinical trials, comprehensive automatic systems of traceability have been used from the Clinical Trial Unit and until administration⁹.

The job of labelling analysis is an exclusive responsibility of regulatory agencies before the distribution of the experimental medications. We healthcare professionals have the duty to confirm that the regulations have been fulfilled, and to ensure that the label contains enough quantity and quality of information to reduce to the minimum any errors in dispensing and administering these medications^10,11. According to the British Royal Pharmaceutical Society¹², Hospital Pharmacy Units must ensure that the research sample has enough quality before its initial dispensing. Therefore, it is the duty of all healthcare professionals involved in clinical trials to report to the regulating agencies any labelling with poor quality, even though it meets current regulations.

Though the mean quality of all samples studied has been high, with proportions over 90% in almost all variables, it is surprising to find that some samples lack the essential information for any medication. For example, 7.4% of samples did not include the clinical trial code, or included several different codes, thus causing confusion among the healthcare staff in charge of dispensing similar samples for different clinical trials. A 5.6% did not specify the conditions for drug storage, and therefore patients at home were not sure whether they should keep the medication in the fridge or at room temperature. Besides, 1% did not include something so basic and necessary to ensure the efficacy and safety of a drug as its expiry date. Even though the lack of these items occurred only in a minority of cases, it represents a major source of errors which might reach the patient.

Management through IVRS/IWRS is one of the exceptions based on which the pharmaceutical companies might be not required to provide all the information collected in Appendix 13; so we held the a priori hypothesis that these labels would have inferior quality. But on the contrary, these labels present a higher mean quality in statistical terms. However, a trend can be observed, which is not statistically significant, towards the lack of the most important items: clinical trial code and period of validity. There seems to be a clear negative relationship, which is statistically significant, between management through IVRS/IWRS and the quantity of information on the label cover. In another sense, we have demonstrated that the cover of labels for blinded samples presents lower quality. This fact can represent a higher likelihood of errors in dispensing; and therefore, measures should be implemented to avoid them, such as physically separating the blinded samples with different doses.

There are no similar articles published which evaluate the quality of labelling in clinical research samples. In a communication to a Congress by Rabuñal¹³, they studied the adaptation of the information in clinical trial samples to current regulations. In the 204 samples they studied, they obtained very similar proportions to ours for the 11 items included in Appendix 13.

In our study, we found that the promoter variable offered a great variability in the statistical analysis. However, the high number of different promoters made it very difficult to find significant differences among them; it would be necessary to increase the sample size in order to draw conclusions. Likewise, a wide multicenter study would be necessary in order to find significant differences in those trends that have been observed, but have shown no statistical significance.

Regarding the study limitations, regardless of the high number of samples studied, it was not possible to achieve enough statistical power to detect significant differences, due to the low number of events in each variable, and the high number of variables studied. On the other hand, even though a double independent assessment of samples was conducted, there is a degree of subjectivity which must be taken into account.

Summing up, the mean quality of labelling is adequate in the majority of clinical trial samples studied. However, there is an alarming lack of essential information, such as clinical trial code and validity period, which could be a potential source of dispensing or administration errors. A higher involvement by healthcare professionals is required, as well as more studies assessing the quality of labelling and its association with medication errors.

 

Conflict of interests

The authors hereby declare there is no conflict of interests.

 

Bibliography

1. Comisión Europea. Normas de Correcta Fabricación de Medicamentos en Investigación (Monografía de internet). 2003 (Citado 10-08-2015). p. 1-20. Disponible en: http://www.aemps.gob.es/investigacionClinica/medicamentos/docs/anexo-13_EC.pdf        [ Links ]

2. Alfaro ER, Sánchez MI, Desongles T, Santos MD. Análisis modal de fallos y efectos aplicado al área de ensayos clínicos. O.F.I.L. 2015, 25;3:145-153.         [ Links ]

3. Hellier E, Edworthy J, Derbyshire N, Costello A. Considering the impact of medicine label design characteristics on patient safety. Ergonomics. 2006;49(5-6):617-30.         [ Links ]

4. Vredenburgh AG, Zackowitz IB. Drug labeling and its impact on patient safety. Work. 2009;33(2):169-74.         [ Links ]

5. Bohand X, Simon L, Perrier E, Mullot H, Lefeuvre L, Plotton C. Frequency, types, and potential clinical significance of medication-dispensing errors. Clinics (Sao Paulo). 2009;64(1):11—6.         [ Links ]

6. Alexander JH, Levy E, Lawrence J, Hanna M, Waclawski AP, Wang J, et al. Documentation of study medication dispensing in a prospective large randomized clinical trial: experiences from the ARISTOTLE Trial. Am Heart J. 2013;166(3):559-65.         [ Links ]

7. Richards CF, Cannon CP. Reducing medication errors: potential benefits of bolus thrombolytic agents. Acad Emerg Med. 2000;7(11):1285-9.         [ Links ]

8. Sanks RJ. A pharmacy manager's perspective on a serious adverse drug event. Am J Health Syst Pharm. 1999;56(9):907-9.         [ Links ]

9. De Pedro Ambrosio V, Frias Lopez Carme. Seguridad desde la recepción a la administración de citotóxicos y ensayos clínicos mediante un sistema de trazabilidad integral automatizado. Hospital ParcTauli, Sabadell. 58^o Congreso Nacional de la Sociedad Española de Farmacia Hospitalaria y Encuentro Iberoamericano de Farmacéuticos de Hospital. Málaga, Octubre 2013.         [ Links ]

10. Quevedo de Torres A, Pérez Bravo L, Fernández-Fernández A. Participación del farmacéutico de hospital en la investigación clínica. Farm Hosp. 1999;23:24-41.         [ Links ]

11. Generalitat. Conselleria de Sanitat. Estudios Clínicos y Observacionales de Productos Farmacéuticos de la Comunitat Valenciana n^o 07 (Monografía de internet). 2011 (Citado 10-08-2015). p. 1-16. Disponible en: http://www.san.gva.es/documents/152919/170020/RevistaEstudiClinic07.pdf        [ Links ]

12. Royal Pharmaceutical Society of Great Britain. Professional Guidance on Pharmacy Services for Clinical Trials Version 1 (Monografía de internet).2013 (Citado 10-08-2015). p. 1-19. Disponible en: http://www.rpharms.com/support-pdfs/professional-guidance--n-pharmac y-services-for-clinical-trials-141013.pdf        [ Links ]

13. Sociedad Española de Farmacia Hospitalaria. 57 Congreso Nacional de la Sociedad Española de Farmacia Hospitalaria. Comunicaciones científicas (Monografía de internet). 2012 (Citado 10-08-2015). p. 88. Disponible en: http://www.sefh.es/sefhpublicaciones/documentos/57-congreso/Congreso-SEFH-comuni caciones-2012.pdf        [ Links ]

 

 

Correspondence:
Correo electrónico: perez_pabhue@gva.es
(Pablo Pérez Huertas).

Recibido el 10 de agosto de 2015;
aceptado el 25 de enero de 2016.