Communicating with conscious and mechanically ventilated critically ill patients: a systematic review

Background Ventilator-dependent patients in the ICU often experience difficulties with one of the most basic human functions, namely communication, due to intubation. Although various assistive communication tools exist, these are infrequently used in ICU patients. We summarized the current evidence on communication methods with mechanically ventilated patients in the ICU. Secondly, we developed an algorithm for communication with these patients based on current evidence. Methods We performed a systematic review. PubMed, Embase, Cochrane, Cinahl, PsychInfo, and Web of Science databases were systematically searched to November 2015. Studies that reported a communication intervention with conscious nonverbal mechanically ventilated patients in the ICU aged 18 years or older were included. The methodological quality was assessed using the Quality Assessment Tool. Results The search yielded 9883 publications, of which 31 articles, representing 29 different studies, fulfilled the inclusion criteria. The overall methodological quality varied from poor to moderate. We identified four communication intervention types: (1) communication boards were studied in three studies—they improved communication and increased patient satisfaction, but they can be time-consuming and limit the ability to produce novel utterances; (2) two types of specialized talking tracheostomy tubes were assessed in eight studies—audible voicing was achieved in the majority of patients (range 74–100 %), but more studies are needed to facilitate safe and effective use; (3) an electrolarynx improved communication in seven studies—its effectiveness was mainly demonstrated with tracheostomized patients; and (4) “high-tech” augmentative and alternative communication (AAC) devices in nine studies with diverse computerized AAC devices proved to be beneficial communication methods—two studies investigated multiple AAC interventions, and different control devices (e.g., touch-sensitive or eye/blink detection) can be used to ensure that physical limitations do not prevent use of the devices. We developed an algorithm for the assessment and selection of a communication intervention with nonverbal and conscious mechanically intubated patients in the ICU. Conclusions Although evidence is limited, results suggest that most communication methods may be effective in improving patient–healthcare professional communication with mechanically ventilated patients. A combination of methods is advised. We developed an algorithm to standardize the approach for selection of communication techniques. Electronic supplementary material The online version of this article (doi:10.1186/s13054-016-1483-2) contains supplementary material, which is available to authorized users.


Background
Communication with hospitalized patients is essential to improve the quality and safety of health care [1]. Patients in the ICU are often deprived of speech and their ability to communicate, because of intubation. There is a significant relationship between the loss of speech and severe emotional reactions among ICU patients, such as a high level of frustration, stress, anxiety, and depression [2][3][4][5][6][7][8][9][10]. The most commonly used communication methods with critically ill patients, like lip reading, gestures, and head nods [7,8,[11][12][13], are time-consuming, inadequate to meet all communication needs, and frustrating for both patients and nurses [12][13][14][15][16][17][18][19]. Current practice in the ICU is to use less sedation in mechanically ventilated patients, which increases the number of patients potentially able to communicate while mechanically ventilated and awake [2,20]. Even though there are numerous alternative methods of communication available and about 50 % of the ICU patients could potentially be served by simple assistive communication tools [21], caregivers currently make little to no use of the devices for patients in the ICU [7,8,22].
Improving communication could be achieved by using a communication algorithm to standardize the approach of selecting various augmentative and alternative communication (AAC) methods. AAC refers to all forms of communication, other than oral speech, that are used to express messages [23]. Although considering the aformentioned importance of adequate communication for the quality of patient care and well-being, to our knowledge and based on an extensive search of websites for societies of intensive care medicine (Dutch Society of Intensive Care Medicine; European Society of Intensive Care Medicine; World Federation of Societies of Intensive and Critical Care Medicine; Society of Critical Care Medicine; Intensive Care Society) and critical care nursing (American Association of Critical-Care Nurses; European Federation of Critical Care Nursing Associations; Canadian Association of Critical Care Nurses; Australian College of Critical Care Nurses), no protocols or guidelines about communication with intubated patients in the ICU currently exist.
The aim of our systematic review was to summarize the current published evidence on communication methods used with adult nonverbal mechanically ventilated patients in the ICU. Our secondary objective was to develop an algorithm for a structured approach of assistive communication devices with mechanically ventilated patients.

Methods
Search strategy, data sources, and study selection A systematic electronic search was conducted in PubMed, Embase, Cochrane, Cinahl, PsychInfo, and Web of Science databases (to November 2015). The screening of titles and abstracts was done by two independent reviewers. Additional file 1 shows our detailed search strategy. We hand-searched the reference lists of retrieved papers for additional studies. Furthermore, we screened the Internet for the registered clinical studies (clinicaltrials.gov). We included all randomized clinical trials, quasi-experimental studies, and observational studies published in English. The population under consideration included all adult patients in the ICU who were conscious with an oral tube or tracheostomy with inflated cuff. Because we aimed for patients who were completely ventilator dependent, we excluded patients who could tolerate cuff deflation. The interventions that are used when the cuff is deflated were therefore excluded (e.g., one-way speaking valves). The main focus of the studies had to be the use of communication techniques between healthcare professionals and patients, not the information content. We adapted our search strategy during the search process, in close collaboration with a medicine literature search specialist of the Free University Medical Library, and allowed any control group and outcome measure due to the minimal research on the subject. Published conference abstracts with no full article were excluded. Identified titles and abstracts were screened against the inclusion criteria to find potentially relevant papers, and subsequently the full texts were reviewed against the eligibility criteria already described. Any disagreements about study selection were resolved during consensus meetings.

Data extraction and quality assessment
A data collection form was used to abstract the data from included articles and assess the study quality. Both the abstracted data and the study quality were checked by a second reviewer. To assess the methodological quality and risk of bias of individual studies, we used the Quality Assessment Tool (QATSDD). This tool can be applied to a diverse range of study designs, including qualitative and quantitative methods, and has clear defined scales. The QATSDD consists of 16 criteria, which all apply to mixed-methods studies. There are also 14 criteria that apply to qualitative studies and 14 criteria that apply to quantitative studies. Test-retest and inter-rater reliabilities have been assessed and ranged from good to substantial (kranging from 0.698 to 0.901) [24].

Data synthesis and analysis
We used the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) checklist for reporting systematic reviews [25]. Because of clinical heterogeneity in the methodology, interventions, and outcome of the included studies, pooling of quantitative data to perform a meta-analysis was not possible and therefore a narrative synthesis was undertaken. To guide our conduct of narrative synthesis we consulted the "Guidance on the Conduct of Narrative Synthesis in Systematic Reviews" of Popay et al. [26].
To accomplish our secondary aim, the studies included in this systematic review were analyzed to detect patient characteristics associated with the use of a specific communication method. The algorithm was developed based on the associations found between characteristics and communication methods, and an algorithm published by Williams in 1992 [27]. During the construct, the algorithm was discussed in a local working group on communication with critically ill patients, consisting of intensivists, critical nurses, and a PhD student.

Quality assessment
Using the appraisal outlined, the overall methodological quality varied from poor to moderate; score 9-35 out of 42, median 17 (see Additional file 2).
Overall the studies were small, with only two reporting an a-priori sample size calculation [12,29]. Only six studies used measurement tools with statistical assessment of reliability and validity, the majority used a subjective assessment. Only four studies used comparator interventions [12,[28][29][30]. All studies had significant limitations in their design. There was a moderate to high risk of bias in the studies included in this review.

Communication boards
A communication board for intubated patients consists of icons and pictures representing basic needs. This was used with three studies, one retrospective cohort and two quasi-experimental studies [12,17,30]. The first study, by Stovsky et al. (1988) [30], stated that a planned communication with a picture board (comprised of 22 pictures with words) increased patient satisfaction, measured with the visual analog scale on satisfaction with communication, in the early postoperative intubation period after cardiac surgery (p = 0.05). Of the patients, 70 % (n = 14) asked for items not indicated on the board [30]. The other two studies used a two-sided board with the alphabet, a picture of the human body, and a pain scale combined with sentences [17] or illustrations [12]. In the retrospective study by Patak [12,17,30].

Speaking tracheostomy tube with inflated cuff
Currently there are six types of specialized talking tracheostomy tubes available to allow communication with an inflated cuff [49]. The Portex Trach-Talk [36,39,42] and Communi-Trach I [42,56] were used with four case series. The Portex BLUSA Tracheostomy Tube was reported in one article with four case reports [49]. These tracheostomy tubes have an additional lumen above the cuff through which air can flow into the larynx to facilitate verbal communication. Intelligible speech, measured by a subjective assessment of improvement of speech, was created in 100 % (n = 20) [39] and 74 % (n = 14) [36] using the Portex Trach Talk, in 90 % (n = 18) with the Communi-Trach I [56], and in 79 % (n = 15) using the Portex Trach-Talk and the Communi-Trach I (both or one of them were used) [42]. With the BLUSA tube, all patients (n = 4) achieved adequate phonation [49]. It took an average of 2.1 days (Portex "Talk") [39] and 5.6 days (Communi-Trach I) [56] before adequate voice intensity for intelligible speech was produced.
The second option, reported in two case series, is the Blom Tracheostomy Tube which incorporates two separate valve mechanisms, through which all of the ventilator-delivered inspiratory air is directed to the lungs and the expiratory air can escape via fenestrations to the upper airway to allow phonation [37,40]. With the Blom Tracheostomy tube, 90 % (n = 9) [37], measured with a subjective assessment, and 100 % (n = 23) [40], assessed with the Assessment of Intelligibility of Dysarthric Speakers, were able to achieve intelligible speech. In the study by Leder et al. (2013) [40], the time to audible voicing was 6.60 min. In both studies some of the subjects were also able to converse over the telephone [37,40]. In the study by Kunduk et al. (2010) [37], two subjects (20 %) experienced clinically important oxygen saturation decreases (<90 %). The other study with the Blom Tracheostomy Tube showed no significant differences (p > 0.05) in oxygen saturation [40].

Electrolarynx
The EL is a battery-powered handheld device which is pressed onto the skin of the neck to transmit the vibrated electronic sound into the oropharyngeal cavity, where the user modulates it to create speech via articulation [57]. We retrieved seven studies on usage of the EL in the ICU setting, of which five were case series [31,33,43,44,46] and two were case reports [47,50]: 72 % (n = 57) of all subjects had a tracheostomy. Results showed successful communication, by creating intelligible speech, with the EL in patients with a tracheostomy in all of the case series: 86 % (n = 19) [31], 100 % (n = 8) [33], 80 % (n = 5) [43], 100 % (n = 2) [44], and 79 % (n = 15) [46]. Creating intelligible speech with the EL in intubated patients was observed for 50 % (n = 2)  [46] and 46 % (n = 6) [44]. One study reported the use of an EL with a nasotracheal tube, of which 50 % (n = 2) had good results, meaning good value of the EL as a means of communication [46]. All but one of the studies measured the success of the EL with a subjective assessment of improved communication. Tuinman et al. (2015) used the self-developed Electrolarynx Effectivity Score. Both case reports reported successful use of the EL in creating intelligible speech with an intubated [47] and tracheostomized [50] patient. In the study by Ewing (1975) [33], the EL was the most preferred by the patients (n = 8) and staff (n = 32) over other available basic communication methods (lip movement, sign language, and writing). In the case series of Summers (1973) [43], clear intelligible speech was produced after 15-30 min of instruction in 60 % (n = 3) and 1-2 hours in 40 % (n = 2) of patients.

High-tech communication intervention
All electronic AAC devices described in the nine studies had common topics about basic communication needs in the ICU on the main screens (e.g., emergency, pain, and emotions) [19,32,34,35,41,45,[51][52][53][54]. Two case series reported the use of voice output communication aids (VOCAs), which are a subset of handheld AAC devices with which patients touch a word-picture icon on a keypad to produce a prerecorded voice message [34,35]. VOCAs were used in 17 % [35] and 27 % [34] of observed communication events, and in both studies all patients were able to successfully generate valid messages. In 70 % [34] and 94 % [35] of the observed communication events, more than one method of communication was applied (e.g., gesture, mouthing words, head nods, and writing).
The computerized AAC devices are specialized computers that contain a database of prestored phrases or pictorials. The selected phrases are voiced by a speech synthesizer [19,32,41,[51][52][53]. The LiveVoice computer of the pilot study by Miglietta (2004) [53] uses various control devices for navigation through the menus; infrared eye-blink detector, touch buttons, or a touchsensitive screen. Over 90 % of the patients felt that the system assisted them in obtaining their needs (pain management, hygiene, comfort, and anxiety). Of the hospital staff, 96 % (n = 42) felt that the LiveVoice improved patient care; this was not further specified. The pilot study by Rodriguez (2012) [19] used a multifunctional computer with touch buttons and a touchsensitive screen. Ten patients (91 %) were satisfied with use of the device, measured with the Patient Satisfaction and Usability Instrument, and showed the ability to independently use the device from day 1. The use of the ICU-Talk communication computer was reported as a case series by Etchels et al. (2003) and MacAulay et al.
(2002) [32,54]. Of the patients who remembered using the computer, 16 % (n = 3) found it useful for creating conversations. Of the nurses, 44 % said the ICU-Talk assisted with patient care. The pilot study by Koszalinski et al. (2015) [52] with the Speak for Myself Computer Pad stated that 95 % (n = 19) thought the device was helpful for communication and it decreased frustration levels. The 'intelligent' keyboard of van den Boogaard and van Grunsven (2004) [45] scored more highly on ease of use (63 %) and satisfaction (88 %) compared with the alphabetical letter board. The gaze-controlled system of Maringelli et al. (2013) [41] improved significantly (p < 0.001) the ability to communicate basic needs and necessities, and decreased the level of anxiety remarkable. The eye-tracking device of Garry et al. (2016) [51] gave all of the patients the ability to communicate basic needs and had a positive psychosocial impact.   [28,55] described the use of AAC strategies for 50 temporarily nonspeaking ICU patients. The oral approaches (e.g., EL, speaking valve) and fine motor approaches (e.g., communication board) were the most recommended techniques. The most successfully served patients were those who were able to use several approaches simultaneously (70-82 % communication

Patient characteristics and barriers for usage of the communication intervention
Patients using the communication boards were orientated without changes in mental status, able to see well enough to read the prints, and had no linguistic problems [12,17,30]. Two out of three studies used orally intubated patients after cardiac surgery [12,30]. Difficulties with the use of the different communication boards were that the board contained too much information [17]. Certain needs and requirements of the patients were lacking on the board [12]. Also, optimal positioning of the board presented difficulties [30].
All tracheostomized patients in which the talking tracheostomy tubes were used were cognitively intact without upper-airway obstruction and had intact muscular function for articulation. Patients had mixed diagnoses [36-40, 42, 48, 49]. Common causes of malfunctioning of the tube included occlusion of the air vent ports, cuff leaks, and kinking of the airflow line tubing. One case report, also using a tracheostomy tube with additional lumen (Vocalaid), stated that it was inadequate for communication due to fatigue after a few minutes and discomfort [48].
The EL was used in patients with mixed diagnosis, intact cognition, and articulatory function [31,33,43,44,46,47,50]. Difficulties in using the EL were due to trouble of understanding the EL voice in the beginning and the unnatural voice quality. In four of the studies, some patients needed assistance in positioning of the device [31,33,44,50].
High-tech communication interventions were used by patients who were cognitive able to communicate and follow simple commands [19,32,34,35,41,45,[51][52][53]. In six studies patients needed to have some muscle power to use the device [19,32,34,35,45,52]. Gaze-controlled and eye-tracking devices were used in paralyzed or physically limited patients with intact visual acuity [41,51,53]. Primary barriers to using the VOCAs were poor device positioning, deterioration of motor and/or cognitive function, and unfamiliarity of healthcare professionals with the use of the VOCA [34,35]. Barriers to using the computerized AAC devices were fatigue, insufficient muscle power or coordination of the upper extremities, and reduced attention span or sedation [19,32,41,45,[51][52][53].

The algorithm
To construct the algorithm we used the patient characteristics and the results which are presented in this systematic review. As a starting point we used the algorithm published by Williams in 1992 [27]. Through discussion, in the aforementioned working group, a hierarchy of assessment tasks to facilitate assessment and selection of communication methods was determined. The algorithm is shown in Fig. 2.

Discussion
Our search of the literature revealed that relatively little attention in critical care research is given to the improvement of communication possibilities with ICU patients. The major finding of our systematic review on communication interventions for conscious and mechanically ventilated patients in the ICU is that in general all four communication intervention types-communication board, speaking valve, EL, and "high-tech" AAC devices-showed a demonstrable improvement in the patient's ability to communicate; a strategy using a combination of communication methods is advisable. The enhanced communication through these devices may provide even small improvements in these vulnerable patients' severe emotional reactions and contribution to healthcare decisions, thereby having a major impact on well-being.
The strengths of this review include the fact that is the first systematic review on the subject to our knowledge, the inclusion of all currently studied communication interventions for ventilated patients, the use of a robust search strategy to identify all studies on the matter, and the use of a validated quality assessment tool to evaluate the quality of the evidence. A limitation of the current state of evidence is that all of the included studies had poor to moderate methodological quality, which has important implications for the generalizability of the results. For example, there were no randomized controlled trials on the subject. A limitation of our results is that we could not conduct a meta-analysis, so no statements can be made on which communication method has proven to be most effective. Another threat to the validity of our results is publication and language bias. Also, when considering the results of the individual studies, the reader has to keep in mind that all patients were aware of participating in an experiment. It cannot be excluded that the increased patient satisfaction and positive outcomes with the communication tools reported could have been due to the increased attention they were receiving from the researchers and nurses, the so-called Hawthorne effect.
Recent data suggest that about 50 % of the patients in the ICU could be served by simple assistive communication tools [21]. The most straightforward method is the use of communication boards. One case report even presented the successful use of a mouthstick stylus (adjustable antenna) fixed on a mouthpiece with a communication board [48]. However, illustrated communication cannot fully comprehend the needs of patients. The specialized talking tracheostomy tubes can be a solution for tracheostomized patient who cannot tolerate cuff deflation. The Portex Trach-Talk was successfully used in four studies, [36,38,39,42]; however, in another publication by Leder and Astrachan (1989) [56] stomal complications were reported in the form of pressure necrosis and wound extensions at the insertion of the airflow line. The more recently developed Portex BLUSA and the Blom Tracheostomy Tube appear to achieve sustained audible phonation [37,40,49]. Nonetheless, more studies are needed to assess safety and to decide whether they can be used as a primary means of communication. The EL seems to be another elegant device because it is easy to manipulate. Its effectiveness was mainly demonstrated with tracheostomized patients [31,33,43,44,46,50], but it also seems to work with orally intubated patients [44,46,47]. An important advantage of most computerized AAC devices is that they can be equipped with different control devices to ensure that physical limitations do not prevent use of the device. None of the articles used the same type of electronic AAC, which makes it impossible to decide which software has the most potential.  The main characteristics to limit usage of assistive communication materials, that were reported in the included studies, are the motor ability of the upper extremities (if needed to control the device), level of sedation, and cognitive status (cognitive fluctuation or deterioration) during critical illness. Also, reduced sensory status (e.g., lack of glasses or hearing aids) may be a barrier for effective communication.
A contributing factor in maintaining ineffective communication with intubated ICU patients is the absence of a systematic method of using various communication interventions [15,18]. We believe that the use of communication interventions with intubated patients in the ICU needs to be imbedded in a communication strategy defined in a protocol. The strategy described in the protocol should be brief, minimally fatiguing, and immediately beneficial to both the patient and the staff [55,61]. Secondly, ICUs need to be equipped with different low-end and high-end AAC devices to improve the ability to choose the most suitable intervention. Thirdly, healthcare professionals need to be trained in the usage of the various AAC devices. Lastly, it is important that the communication need and success of communication interventions are tracked in the medical chart of the patient. In this way, communication interventions will be tailored to the specific needs of speechless critically ill patients. Therefore, we developed an algorithm with a hierarchy of assessment tasks to facilitate the assessment and selection of a communication intervention with conscious mechanically intubated patients in the ICU. This algorithm is a starting point. Because the evidence for its design is scarce, it needs to be validated and possibly adjusted in clinical practice by health professionals and/or AAC experts. The algorithm could also be used for future studies.
The main question that still needs to be addressed is: what communication intervention works best for which ventilated patient? Further research with larger, multicenter studies is therefore needed to compare the effectiveness of the various communication techniques as well as introducing new innovative communications techniques. Future research should include specifics regarding baseline patient data, prerequired patient characteristics necessary to use the devices, level of sedation, training duration of staff and patients for usage of the communication device, and costs. Also, the use of different communication strategies needs to be studied.

Conclusions
A summary of current available research on the various communication methods available for ventilated patients in the ICU is presented in this systematic review. The results of the four presented communication intervention types (communication board, speaking valve, EL, and "high-tech" AAC) all showed an improvement in the communication with mechanically ventilated patients. A combination of various methods may create the most effective communication option. However, the results should be interpreted with caution because evidence on the matter is limited and most studies did not have a comparator. Limitations in the use of assistive communication tools are the level of sedation, decreased cognitive function, and muscle power. We developed an algorithm to guide both clinical practice and further research for the use of assistive communication with intubated patients in the ICU.