This study, covering an overview of the practice of ultrasound, was carried out in 47 health facilities, including the two Teaching Hospitals, which are the reference centers of the country. This is the first study carried out in this framework in our country. It will serve as a repository of data on this subject. We have made a reasoned choice based on a sample of health centers in Lomé. Such a choice may constitute a limit for this study since it does not take into account all the health centers in Lomé. However, given the selection criteria for this sample, we consider that it is validly representative of the health centers of the city of Lomé.

Ultrasound equipment was usually acquired in new condition. This can be explained by the appearance on the Togolese market of Chinese branded ultrasound scanners which are at very affordable prices compared to ultrasound scanners of European and American origin.

Our study revealed that all ultrasound machines had a convex probe, while only 37.9% of ultrasound scanners had a linear probe. This finding differs little from that of Girisgin et al. in Turkey who found that all ultrasound scanners

N = number; % = percentage.

*Other: surface organs and soft tissues.

used in the bedside in emergency services had both convex and linear probes [3] . Our result could be explained by the fact that the most extensive explorations remained those of the deep organs (liver, kidneys, pancreas, uterus, ovaries, etc.) considered to be part of general ultrasound. It is for this reason that the WHO since the years 96 recommended that for the purchase of a general purpose ultrasound system, the standard probe should have a frequency of 3.5 MHz [4] .

Radiologists were the ones who carried out the majority of the ultrasound examinations in the public centers while in private they accounted for just under half of the performers. Physicians enrolled in radiology specialties played a significant role in these private structures. The European Society of Radiology in 2013 found a mixed result [5] . There were hospitals in which the majority of ultrasound examinations were carried out by radiologists whereas it was not the case in other hospitals. In our study, a radiologist alone was obliged to cover 3 to 4 ultrasound centers since until 2011, Togo had only 9 radiologists, 3 of whom were exclusively in private practice [6] . The place of doctors in radiology specialization in this practice deserves to be seriously discussed in the reflections to come to define exactly their exercise. No general practitioner or paramedic health worker performed ultrasound in all centers visited while this competency is transferred to other health professionals in other countries of Africa. This is the case for example in Senegal where general practitioners and midwives are trained in ultrasound [7] . The need to train non-radiologists in this practice in our country is therefore no longer to be demonstrated. The training of a sonographer is long: one year of minimum training for a specialty (cardiovascular, abdomen, pelvis, fetus...). In addition, for ultrasound diagnostics, the sonographer often needs the advice of another colleague who, in most cases, does not work in the same health facility or in the same city as himself, the tele-ultrasound could appear as an alternative [8] .

In all the visited centers, the ultrasound report was made using pre-established records. This mode allows faster editing. It is adapted to a given pathology or to an organ. This is often the case for obstetric ultrasound reports [9] . It is often handwritten in our context and does not require a secretariat. The disadvantage is that it is not often adapted to the question posed by the prescriber. Indeed, the radiological report is a written transcription, in clear and unambiguous terms, and transmission, of the different stages of the radiological examination: indications, techniques of realization, results, synthesis and medical conclusion [10] . Since ultrasound is a “dependent operator”, the clinician therefore needs quality criteria to know whether he can trust the ultrasound he has in his hands or not. The length can not be standardized and should be adapted to the pathology to be described and the question asked.