LaComm, A French Communication Analysis Software (1/3)

ABSTRACT There is an increasing awareness that effective communication skills of health care professionals (HCP) are needed to achieve desired health outcomes. Consequently, there has been a growing demand for effective methods to assess communication skills and communication skills training programs. A software –for the French language– has been developed for these purposes. This software is called “LaComm” (Logiciel d’Analyse de la Communication). The aim of this software is to analyse the verbal communication. This software identifies communication skills used by professionals as well as their effects on patients and relatives utterances. LaComm analyses utterances transcripted from audiotaped interviews. LaComm was developed to assess HCP communication skills (by an heuristic method) and to assess topics discussed in HCP-patient interviews (by a topics analysis method).


Preserving the best possible quality of life for cancer patients and their families has become a major goal in cancer care. However, the cumulative effects of stressors related to cancer care, many of which involve communicating with patients and relatives, may lead to the development of health care professionals burnout and poor quality of care. Many health care professionals lack the psychosocial knowledge and communications skills needed to identify patients’concerns and problems. Teaching strategies are therefore being developped to improve health care professionals’communication skills. Available tools for analysing two persons interviews require highly trained psychologist to achieve inter-rater reliability. Theses scoring systems are also time consuming. Moreover, none of available tools allows the analysis of three persons interviews. More empirically, it has been shown in a randomised design that nurses which participate to a communication skills training program increased use of emotional words in their interactions with patients (Razavi et al, 2002). For all these reasons, a software designed specifically for assessing health care professionals communication skills was considered as needed and therefore developed. This software –developed for the French language– is called “LaComm” (Logiciel d’analyse de la Communication).


METHODS LaComm is constructed on the basis of an empirical approach referring to various linguistic models. Each utterance is as a “chain of Markov” (1913), which respects the rules of grammar (Transformational Generative Grammar formalized by the Augmented Transition Network Grammar) (Chomsky, 1957, 1965; Woods, 1970). It should be recalled also that each utterance which is analysed by the Lacomm appears in the context of very specific speech situations (communication in medicine in general and oncology in particular) (Austin, 1962; Searle, 1969). For these reasons, each utterance may be tagged by the LaComm with relatively small databases of “preset categories of communication skills”. The “preset categories of communication skills” were adapted from the categories of CRCWEM (Booth, 1991) and were redefined by a panel of experts. LaComm uses databases of «preset categories of communication skills» built on pragmatical knowledges related to health care professionals clinical practice, on psycho-oncological knowledges, on emotional literature knowledges, on dictionary derived knowledges, and on empirical knowledges derived from actual and simulated interviews performed by physicians. LaComm has greatly benefited from databases of an empirical study (Razavi et al., 2003, Delvaux et al., 2005,). To generate each database of “preset categories of communication skills”, canonical utterances were extracted and widened by using dictionaries of synonyms. All canonical utterances were validated by a panel of experts. To analyse each utterance, LaComm uses a heuristic method of matching at two level. The first level can be described as follow: each utterance is examined by all “preset categories of communication skills” databases until matching. The second level can be described as follow: each category is defined by a strict combination of databases stored utterances. LaComm may provide also another type of content analysis: each utterance is examined by all words or expressions databases of various topics (psychological, medical and others). LaComm calculate the number of utterances tagged by each “preset categories of communication skills” and by each topics databases. Lacomm is written in Visual Basic 6 and use Microsoft.Jet.OLEDB.4.0 to link to ACCESS 2000.

1. Assessment skills and «preset categories of assessment skills» databases.
Assessment skills allow HCP to assess patient physical problems or concerns, emotional expressions or cognitions. Assessment skills include open assessment, half-open assessment or closed assessment skills. “Preset categories of assessment skills” databases were developed for each of these categories of communication skills.

1.1. Open assessment skills and «preset categories of open assessment skills» databases.
Open assessment skills allow HCP to assess a wide range of concerns, problems or subjective feelings.
«Preset categories of open assessment skills» databases include categories of utterances which are introductive utterances (How are you doing? Tell me) or articulative utterances (Something more to be added?).

1.2. Half-open assessment skills and «preset categories of half-open assessment skills» databases.
Half-open assessment skills allow HCP to assess more precisely patient concerns; problems or subjective feelings. Half-open assessment skills include clarificative half-open assessment skills and coreferenced half-open assessment skills.
1.2.1. Clarificative half-open assessment skills allow HCP to assess more precisely issues previously assessed or not.“Preset categories of clarificative half-open assessment skills” databases include categories of short utterances (To summarize?), precise utterances (What do you mean?) or affirmative utterances (Tell me what occurred since the last treatment).
1.2.2. Coreferenced half-open assessment skills allow HCP to assess issues related to one or more already explicitly evoked topic. “Preset categories of coreferenced half-open assessment skills» databases include categories of coreferenced utterances (What do think you about it?), precise coreferenced utterances (What do you feel about of your treatment?) or uncoreferenced utterances (What do you think?).

1.3. Closed assessment skills and «preset categories of closed assessment skills» databases.
Closed assessment skills allow HCP to assess precisely a dimension. “Preset categories of closed assessment skills” databases include categories of utterances which lead patient to give a short answer with a yes or a no (Did you begin the treatment?) or to give a precise answer (How many days?).“Preset categories of closed assessment skills” databases include also categories of utterances which are related to the interview context (Should I explain?), short utterances (Did you see that?) or suggestive utterances (You do not have pain, isn't it?).

1.4. Checking skills and «preset categories of checking skills» databases
Checking skills allow HCP to check the understanding “Preset categories of checking skills” databases include categories of utterances which check if HCP has correctly perceived/understood what the patient expressed (Really?), or if the patient has correctly perceived/understood what HCP expressed (Do you understand what I want to say?).

next >>>>

ACKNOWLEDGEMENT This research program was supported by the «Gouvernement de Bruxelles-Capitale » of Belgium