Our survey targeted the nurses and midwives practicing in 2017 in any of the hospital services where drugs were prescribed. From this source population, we selected from the administrative list of the paramedical staff at the CHU of Cocody a sample based on the oral consent and the physical presence. Indeed, periodic visits by one of physicians of pharmacology department according to an appointment calendars allowed, after the explanation of the questionnaire, to recruit the participants in their workplace. Thus, all nurses and midwives present during the study period regardless of their age, gender, nationality, year of professional experience as long as they have agreed to participate in the survey were included in the study. On the contrary, the nurses and the midwives employed in the selected services but absent (with the exception of those on night shifts) at the time of the survey or who did not give their consent or who did not complete the survey form were not included in the study.

In a first step, we developed based on the literature reviews [5] [6] [7] [8], a standardized and anonymized questionnaire containing in the majority of cases open-ended (closed) questions. The questionnaire was designed on the classical model for assessing attitudes, knowledge and practice. It was, after validation at the Clinical Pharmacology Department, pre-tested in the Medical and Gynecological emergency services for improvement. Thereafter, the registrar of the Clinical Pharmacology Department made several rounds in the selected services to interview the nurses and midwives who were previously informed about the objectives of the study for the completion of the questionnaire. In some cases (lack of time, heavy workload, absence due to night shift, etc.), the questionnaires were given to the nurses and midwives for self-administration and then collected the following day. The perception of pharmacovigilance was evaluated from 13 questions grouped in 5 broad variables. It was:

· The professional characteristics (profession, name of the specific hospital service, and seniority in the position).

· General knowledge on pharmacovigilance (source of information, pharmacovigilance teaching hours in the basic training, existence and location of a pharmacovigilance unit at the CHU of Cocody, importance given to pharmacovigilance in practice).

· Reporting of adverse drug reactions (existence of ADRs in the service, types of adverse reactions reported, reporting methods, reasons for non-reporting).

· The expectations of nurses and midwives after reporting an ADR.

· Measures to improve the rate of reporting of ADR by the nurses and midwives.

The data was analyzed using the EPI info7.4 software for the capturing of the information and the S.P.S.S. program. 17.0 (Statistical Package for the Social Sciences version 17.0) for statistical analysis. A Pearson Chi-square test was used to compare the results with a significance level of 5%.

When collecting the data, we respected the anonymity of the nurses and midwives and ensured that their activities were not disrupted during working hours. Moreover, we obtained the oral agreement of the Head of the services and the Unit Care Supervisors (UCS) of the selected services. In addition, a written authorization from the Director of the Medical and Scientific Department of the University Teaching Hospital of Cocody was obtained before the beginning of the study. Finally, no healthcare provider completed the questionnaire under coercion.

Of the 272 paramedics (196 nurses and 76 midwives) surveyed, only 39.33% (86 nurses and 21 midwives) voluntarily answered the questions (Table 1). They mainly worked in the service of internal Medicine (58.46%) and Surgery (52.11%) and had in the majority of cases (50.6%) a professional experience of less than 10 years (Table 2). The other primary information such as age, sex, specialty and place of residence were incorrectly entered on the questionnaire and could not be showed in a table.

In our study, 57.94% (62/107) of paramedics (52 nurses and 10 midwives) had already heard of pharmacovigilance (Table 2), mainly during their basic training (43/107). However, only 5.60% (6/107) of these respondents were satisfied with the volume of hours allocated to the pharmacovigilance module (Table 2). In addition, 19.63% (21/107) of respondents were aware of the existence of a pharmacovigilance unit (Table 2) that they had pointed to be inside the pharmacy (11/21) of the University Teaching Hospital of Cocody or at the Clinical Pharmacology

*Emergency service: Emergencies of Medicine, Surgery, Gynecology and Pediatrics; **Medicine: Pneumophtisiology, intensive care unit, Rheumatology, Gastroenterology, Neurology; ***Surgery: Traumatology, Digestive Surgery, Urology, Ophthalmology, Oto-Rhino-Laryngology, Stomatology, Pediatric Surgery.

Department (10/21). Furthermore, the paramedics (75/107) found pharmacovigilance important (21.49%) or very important (48.59%) in their daily practice (Table 2).

In our survey, 40.18% (43/107) of the paramedics had already come across an adverse event (Table 3), mainly an adverse drug reactions (32/43). They reported these events in 76.74% (33/43) of the cases to a hierarchical supervisor

(20/33). The main causes of non-reporting of adverse drug reactions (Table 3) were related to their frequent occurence (4/10) or their benign nature (2/10).

Specific instructions such a code of conduct (37.38%) and a feedback after reporting (28.97%) were the main expectations of our respondents (Table 4). They also wanted a regular visit of the pharmacovigilance monitors in their services (34.57%) and the provision of the reporting forms to stimulate a spontaneous reporting of adverse drug reactions (Table 4).

In our study, the response rate of the paramedics varied significantly according to the profession (nursing or midwifery) (Chi² = 6.058, ddl = 1, P < 0.02) and the service in which they are working (Chi² = 18.634, ddl = 4, P < 0.001). This could explain the various level of their knowledge on pharmacovigilance. In addition, most of our paramedics (54/107) had a professional seniority of less than 10 years. This was not reported in several studies [5] [6] [7] [8] [9] in which the majority of nurses interviewed had more than 11 years of working experience.

In our survey, 57.94% (62/107) of paramedics had already heard of the term pharmacovigilance mainly during their basic training (Table 2). However, the pharmacovigilance teaching hours considered insufficient and reported by 94.39% (101/107) of the paramedics reflects the need for a continuous training. There was no significantly different relationship between having previously heard about pharmacovigilance and nurse or midwife status (p = 0.14) and hospital service (p = 0.124). On the other hand, there was a significant difference between having previously heard about pharmacovigilance and professional experience (p = 0.028). This professional experience was not statistically related to declaring pharmacovigilance as an important activity (p = 0.61).

Apart from that, 80.37% (86/107) of the paramedics in our study were unaware of the existence of the pharmacovigilance unit and wrongly located it at the pharmacy (11/21). This could be a barrier to reporting of the adverse drug reactions at the CHU of Cocody. Our results were comparable to those of a Saudi study [10] in which 62.5% of healthcare professionals said they did not know the term “pharmacovigilance”. Furthermore, in two (2) Italian and Portuguese studies [6] [7], the nurses did not know the existence of the national pharmacovigilance system in 50% and 58.1% of cases respectively.

In our work, the ignorance of the reporting procedures was another factor that could negatively influence the perception on pharmacovigilance. In fact, the majority of the 33 respondent paramedics (n = 20, 60.60%) had reported the adverse drug reactions to their hierarchical supervisor and none of these events were reported to the pharmacovigilance unit. This did not comply with the Regulatory Act that creates the national pharmacovigilance system in Côte d’Ivoire [2]. Under these conditions, a special emphasis should be made on the procedural aspects of the Act if the desire is to increase the number of adverse drug reactions reporting at the CHU of Cocody. Furthermore, in our survey, 4 out of 10 respondent paramedics did not report the adverse drug reactions because of their frequent occurrence (Table 3). This was not justified because all adverse drug reactions should have been reported, especially those that are unexpected or serious [2]. In an Italian study [6], the poor knowledge of the pharmacovigilance system, the fear of conflicts with doctors and the fear of legal reprisals were the main reasons for the under-reporting of the adverse drug reactions by nurses.

In our study, paramedics mainly expected a precise code of conduct and a feedback on a reported event (Table 4). In addition, to stimulate a spontaneous notification, most of the respondent paramedics had suggested a regular visit of the pharmacovigilance monitors in their different services (34.57%, n = 37) and the provision of the reporting forms (30.84%; n = 33) as shown in Table 4. These expectations and measures expressed by the paramedics show their interest to see the pharmacovigilance been practiced at the CHU of Cocody. This interest should be taken into account in the organization of awareness and continuous training in the practice of pharmacovigilance in Côte d’Ivoire (Ivory Coast). In fact, regular visits by pharmacovigilance monitors to hospital services could help staff paramedic to detect adverse drug reactions, to identify and resolve problems encountered by nurses and midwives in process of declaration of adverse drugs reaction and develop a special relationship with them. This attitude, coupled with the disponibility of the declaration forms, could positively influence the perception of pharmacovigilance by paramedical staff. Overwise, Feedback with precise management is an important device in pharmacovigilance. This feedback, which also mentions the responsibility or not of the suspected drug (s), should be included in the patient file and thus testify to the occurrence of the adverse drugs reaction.