Medication mistakes are one of the leading causes of morbidity and mortality in critically ill patients. According to the SYREC trial—conducted in 79 Intensive Care Units (ICU)—in Spain medication mistakes were the most common harmless incidents (31.2%), and cause for 11.6% of adverse events recorded.1 In studies conducted in the United States it has been reported that medication mistakes in these patients are more common compared to other hospitalized patients with a 2–3 times higher risk of adverse events and a 2.5 higher mortality rate.2,3 Several factors contribute to the higher risk of adverse events preventable with medication at the ICU setting, among them: the greater severity of critically ill paients and the complexity of treatments they need: the use of numerous drugs—many high-risk drugs—and the IV administration of a large number of drugs that require dose estimation, and are often administered via continuous IV perfusion.4,5

To improve the safety of drugs at the ICU setting a multimodal and multidisciplinary approach4,5 is needed to facilitate the implementation of specific mistake prevention practices in all the processes that determine the safety of such practices. Proactive assessment tools have proven very effective to help health centes analyze their processes, identify areas of risk, and prioritize practices for improvement that should be implemented. In the United States, the Institute for Safe Medication Practices (ISMP) developed the ISMP Medication Safety Self Assessment for Hospitals,6 a self-evaluation survery to conduct a comprehensive and extensive analysis of the safety of drugs administered at the hospital setting that has been used in different countries now.7–11 With support and funding from the Spanish Ministry of Health and the technical collaboration of an expert working group from different hospitals, the ISMP-España adapted this survey to the Spanish healthcare practice. Therefore, in 2007 the «ISMP Medication Safety Self Assessment for Hospitals»12 was published. It was updated a years later.13 From this latest iteration of this document, ISMP-España, the Intensive Medicine and Critical Care Pharmacist (FarMIC) working group of the Spanish Society of Hospital Pharmacists (SEFH) and the Spanish Society of Intensive and Critical Care Medicine and Coronary Units (SEMICYUC) drafted the «ISMP Medication Safety Self Assessment for Intensive Care Units»14 to have a specific tool that ICUs could use to make detailed assessments of drug safety in this specific setting, identify the critical areas of risk, and with this information, plan safe practices that should be implemented at local level in each center to minimize the risk of making mistakes.

Once the survey for ICU was published, SEMICYUC, SEFH, and ISMP-España decided to conduct a nationwide study to know the degree of implementation of safe drug practices at the ICU setting and identify any opportunities for improvement in which collaborative effort is required to improve the safety of critically ill patients.

Descriptive multicenter study of the degree of involvement of the safety practices included in the «ISMP Medication Safety Self Assessment for Intensive Care Units».14 It was based on the self-assesment conducted by ICUs and Hospital Pharmacy Services that volunteered to participate in the study and completed the questionnaire from March to September 2020.

The study was published on SEMICYUC and SEFH website and e-mail lists. To register responses to the survey a computer application installed in the ISMP-España website was used to guarantee the confidentialiy of the information provided.

A total of 39 ICUs from 12 different Spanish autonomous communities participated in the study plus ICUs from Andorra. The characteristics of these ICUs are shown on Table 1. A total of 20% of these ICUs belonged to hospitals with 99–199 beds, 37.5% to hospitals with 200–499 beds, and 42.5% to large hospitals with ≥500 beds. Regarding the type of ICU included in the study, most were polyvalent units and only 3 other type of units (2 doctors and 1 coronary artery disease specialist).

Table 2 shows the overall results obtained for the survey in the overall 40 ICUs and in the different groups of ICUs established based on characteristics of functional dependency, size of the hospital, type of ICU, and teaching activity. The mean score to the overall survey in all ICUs was 436.8 points, which is equivalent to 49.2% of the theoretical maximum value observing a wide range of values (20.9%–82.3%). When rates on the maximum possible value observed were compared in the different groups of ICU, no statistically significant differences were found among the established groups (P > .05).

Fig. 1 shows the mean rates on the maximum possible value obtained for the 10 key elements in the overall number of ICUs, which graphically shows areas of a higher risk. Elements iii and viii on the addition of pharmacy specialists to ICUs and on the competence and training of healthcare workers in drugs and safety practices showed the lowest rates of all (31.2% and 33.2%, respectively). The key elements on the availability and accessibility to information on patients and drugs (element i), standardization, storage, and distribution of drugs (element v), and quality and risk management programs (element x) also showed values <50% (44.0%; 43.4%, and 46.7%, respectively). The remaining key elements showed rates above 50%.

Figure 1.

Results obtained in the overall amount of Intensive Care Units (ICU) (N = 40) for the 10 key elements and for the complete survey expressed as rates on the maximum possible value.

Abridged description of key elements: I. Availability and accessibility to information on patients and drugs. II. Communication of prescriptions and other type of information on medication. III. Adition of pharmacy specialists to ICUs. IV. Labeling, packaging, and name of drugs. V. Standardization, storage, and distribution of drugs. VI. Acquisition, use, and follow-up of the different devices for the administration of drugs. VII. Environmental factors and human resources. VIII. Competence and training of healthcare workers in drugs and safety practices. IX. Education to patients and families. X. Quality and risk management programs.

The analysis of the results obtained for each evaluation item provides information at a more detailed level that can be useful to know the degree of implementation of specific practices and be able to identify those that should be approached prioritarily to improve safety. A thorough analysis of the 147 evaluation items discussed would exceed the length of the article. Tables 3 to 5 show the values of several items on the addition of new technologies to healthcare continuity and use of high-risk drugs. The overall results can bee read in a report published by SEFH, SEMICYUC, and ISMP-España.17

On the implementation of technologies that impact the safety of drugs with the corresponding elevated maximum values in the survey (Table 3), low rates of 37.5% were seen for item #6 on electronic prescriptions integrated into the hospital information systems while 43.3% rates were reported for item #30a on the availability of systems to support the clinical decision-making process. The use of a code reader in the administration showed rates of 3.4% only while the use of smart infusion pumps to administer high-risk drugs had a higher value of 52.5%.

Regarding the implementation of standard procedures to match patients’ medication and ICU admission when transferred from a different ICU or discharged from the ICU (items #18, #19, and #20), the mean rates on the maximum possible value obtained were 51.3%, 58.1%, and 71.3%, respectively (Table 4).

Several items associated with the safe use of high-risk drugs had low implementation rates (Table 5). This is the case of item #82 (11.7%) on the elimination of vials or ampoules of electrolyte concentrations from the ICU setting or else be stored separately from other drugs with additional safety measures. Also, item #85 on the storage of neuromuscular blocking drugs (29.4%), and item #69 on the preparation by the Hospital Pharmacy Service of standard IV solutions of high-risk drugs that are not available in the market (25.3%).

Finally, we should mention that item #44 on the availability of a pharmacist assigned to the ICU as part of the healthcare team long enough to perform the clinical activities that the ICU requires, had a mean score of 5.80 ± 5.27 for a maximum value of 16 (36.3%).

Back in 2017, SEMICYUC and SEFH signed a collaboration agreement to lead and support common initiatives in the healthcare, teaching, and research settings.18 Both societies worked on a multidisciplinary basis to establish the safe use of drugs in critically ill patients, recommendations, and clinical practice guidelines. Also, to delve into the epidemiology of mistakes and drug-related adverse events. Drafting the «ISMP Medication Safety Self Assessment for Intensive Care Units», and conducting this study are the early results of this initiative.

This study results provide an overall perspective on the safety of the process of using drugs at the ICU setting and reveal that there is significant room for improvement. Although a wide variety of scores to the overall survey were obtained in the different ICUs of the sample—indicative of the exisence of differences in the degree of implementation of safe practices—the analysis of data revelaed that there are safe and effective practices whose rate of implementation is very low or practically zero in many different ICUs. In our opinion, the information collected is very uself for SEMICYUC and SEFH to prioritize areas where their efforts should be oriented to improve the safety of critically ill patients.

Intensive care unit pharmacists play an important role in the safety profile of the drugs administered at the ICU setting. Evidence supports including them as part of the ICU multidisciplinary team to minimize mistakes, adverse events, and mortality.19,20 Some of the functions associated with intensive care unit pharmacists are helping in the clinical decision-making process, being involved in quality programs to improve drug management, participating in the development, implementation, and follow-up of the ICU pharmacotherapeutic protocol, and implementing new technologies.21 However, in Spain, the involvement of intensive care unit pharmacists in multidisciplinary teams is fairly limited.18,22 The results obtained for key element iii of the survey were the lowest of them all (31.2%), indicative of the importance of the role of intensive care unit pharmacists for SEFH and SEMICYUC.

Element viii on competence and training of healthcare workers regarding drugs and safety practices also had a low degree of implementation despite being one of the fundamental pillars to minimize the risks associated with healthcare.23,24 The use of educational strategies based on clinical simulation and a multidisciplinary approach to reduce medication mistakes at the ICU setting is advised25,26 since teamwork and training have both proven capable of minimizing errors and reducing the mortality rate.25,27

This study demonstrates the need for encouraging the implementation of different technologies to minimize prescription and administration mistakes that are the most common errors at the ICU setting.2–4 Only half of the ICUs have infusion pumps with smat technology available to minimize dose administration mistakes or incorrect velocities. Only 37.5% and 43.3% of ICUs have integrated electronic prescriptions and clinical decision-making support systems, respectively. We should mention that the use of a bar codes in the administration of drugs—considered the mot effective barrier to prevent errors and guarantee traceability—is minimum. In Spain, the development of this technology has always been hindered by the lack of bar code unique identifiers in drugs.

Other safe practices recommended by the World Health Organization in its third global patient safety challenge—“Medication Without Harm”—and by other organizations are aimed at minimizing drug mistakes in healthcare transfares and high-risk drugs,26,28,29 common mistakes at the ICU setting associated with an elevated risk of damaging the patients.30,31 Study data indicates that, little by little, ICUs are establishing practices to guarantee the correct continuity of medication. However, its implementation still needs to be improved. On the safety practices associated with the management of high-risk drugs, it is surprising to see the low rate of implementation of practices considered a priority or emblematic regarding patient safety like the elimination of vials and ampoules of potassium chloride, which can be attributed to the lack of potassium concentration solutions in bags in Spain.

As far as we know, no self-evaluation surveys have ever been conducted specific to the ICU setting or with the same tools, which is why, although desirable, we cannot compare the results of this study to the results of other studies of similar characteristics. However, we should mention that, in the latest study on the implementation of safe practices conducted in 2011 among 165 Spanish hospitals, with the early version of the hospital general survey,32 the lowest values of all were also obtained for the element associated with competence and training of the healthcare personnel in both drugs and safe practices (29.8%). Although the items of both surverys are not the same, there is no doubt both deal with measures considered essential to transform healthcare systems and that these measures have not been implemented in the Spanish healthcare practice like we said before. Other key elements (i, ii, iv, vii, and ix) showed slightly higher rates in this study at the ICU setting probably because, over the years, several safe measures have been implemented in our country. In this sense, we should mention that element vi on drug administration devices showed rates of 61.7%, compared to 46.7% reported in the hospital setting, indicative that significant improvements have been made in this area. Therefore, it has been confirmed that more smart infusion pumps, pumps with free-flow protection and specific systems for the administration of oral solutions and enteral nutrition have been used.

The study has several limitations associated with the methodology used. In the first place, the sample could be non-representative of the activity developed at the ICU setting because it was not randomized. Also, the number of ICUs involved in the study was smaller compared to the number of ICUs initially anticipated since the study was disclosed shortly before the COVID-19 pandemic started, which has impacted ICUs so negatively. That is why different units that wished to participate in the survey could not eventually do so due to the high and continuous care load sustained by the healthcare system. However, the study was closed because its primary endpoint was to identify those areas with highest risk for the ICUs. Also, because the results proved the existence of common problems in most centers, which is why we think that the information obtained from areas with most room for improvement can be generalized.

Other limitations we should mention here are those associated with this type of self-evaluation tools. Therefore, the instructions to conduct the self-evaluation survey indicate that it should be conducted by a multidisciplinary team aware of the reality at the ICU setting. However, noone checked any of this. Also, we should mention that there could have been variability regarding the interpretation of the different items of the survey by the teams of each hospital, which can eventually affect the results. Although the hospitals included in the working group were instructed to confirm that everybody had understood the evaluation items proposed, the reproducibility of the survey was never validated.

In conclusion, this study allowed us to identify numerous safe practices whose degree of implementation at the ICU setting is low but should be dealt with to reduce medication mistakes in critically ill patients. Similarly, we believe that this study promoted the use of self-evaluation surveys at the ICU settings, which contributed to making all healthcare workers familiar with safe practices regarding the use of drugs and then be able to launch actions at local level.

All the undersigned authors contributed to the design and methodology of the study. Also, they directed the survey and contributed to the publication of the study. Dr. Otero and Dr. Martín Muñoz performed the preliminary analysis of data that were reviewed by all the authors. Dr. Otero and Dr. Merino de Cos were involved in the manuscript first draft. All the authors reviewed the results and contributed to the manuscript final version that was latter approved by all the authors.

This study has been funded by SEFH and SEMICYUC.

None whatsoever.