post for guru answers only econ ct3

Z

Q Q

T p o N P Q

Y

1

2

3

S

h 1

. Evid. Fortbild. Qual. Gesundh. wesen (ZEFQ) (2016) 110—111, 26—35

Available online at www.sciencedirect.com

ScienceDirect

journa l homepage: ht tp : / /www.e lsev ier .com/ locate / ze fq

UALITÄT UND SICHERHEIT IN DER GESUNDHEITSVERSORGUNG / UALITY AND SAFETY IN HEALTH CARE

aking up national safety alerts to improve atient safety in hospitals: The perspective f healthcare quality and risk managers ationale Sicherheitshinweise zur Verbesserung der atientensicherheit in Spitälern nutzen: die Perspektive von ualitäts- und Risk-Managern

vonne Pfeiffer1, David Schwappach2,3,∗

ETH Zurich, Department of Management, Economics, and Technology, Zurich, Switzerland Swiss Patient Safety Foundation, Zurich, Switzerland Institute of Social and Preventive Medicine (ISPM), University of Bern, Switzerland

ubmitted/eingegangen 17 August 2015; revised/überarbeitet 2 December 2015; accepted/akzeptiert 11 December 2015

KEYWORDS National event reporting system; learning from national safety alerts; risk management

Abstract Background: National safety alert systems publish relevant information to improve patient safety in hospitals. However, the information has to be transformed into local action to have an effect on patient safety. We studied three research questions: How do Swiss healthcare quality and risk managers (qm/rm1) see their own role in learning from safety alerts issued by the Swiss national voluntary reporting and analysis system? What are their attitudes towards and evaluations of the alerts, and which types of improvement actions were fostered by the safety alerts?

Methods: A survey was developed and applied to Swiss healthcare risk and quality managers, with a response rate of 39 % (n = 116). Descriptive statistics are presented. Results: The qm/rm disseminate and communicate with a broad variety of professional groups about the alerts. While most respondents felt that they should know the alerts and their

∗ Corresponding author: David Schwappach, Swiss Patient Safety Foundation, Asylstr. 77. 8032 Zurich, Switzerland. Tel.: +41 (0)43 244 14 80; Fax: +41 (0)43 244 14 81. E-Mail: [email protected] (D. Schwappach).

1 qm/rm is an abbreviation for ‘‘quality and risk manager’’

ttp://dx.doi.org/10.1016/j.zefq.2015.12.007 865-9217/

Taking up national safety alerts to improve patient safety in hospitals 27

contents, only a part of them felt responsible for driving organizational change based on the recommendations. However, most respondents used safety alerts to back up their own patient safety goals. The alerts were evaluated positively on various dimensions such as usefulness and were considered as standards of good practice by the majority of the respondents. A range of organizational responses was applied, with disseminating information being the most common.

An active role is related to using safety alerts for backing up own patient safety goals. Conclusions: To support an active role of qm/rm in their hospital’s learning from safety alerts, appropriate organizational structures should be developed. Furthermore, they could be given special information or training to act as an information hub on the issues discussed in the alerts.

SCHLÜSSELWÖRTER Nationales Ereignis- berichtssystem; Lernen aus nationalen Sicherheitshinweisen; Risikomanagement

Zusammenfassung Hintergrund: Nationale Fehlermeldesysteme veröffentlichen relevante Information für die Verbesserung der Patientensicherheit in Spitälern. Damit die Warnhinweise einen Effekt auf die Patientensicherheit haben können, muss die enthaltene Information zu lokalen Maßnahmen führen. Wir untersuchten drei Fragestellungen: Wie die Qualitäts- und Risk-Manager (qm/rm2) Schweizer Spitäler ihre eigene Rolle beim Lernen aus Warnhinweisen (,,Quick-Alerts‘‘), die vom freiwilligen Schweizer Berichts- und Analysesystem herausgegeben werden, einschätzen; welche Einstellungen und Bewertungen der Warnhinweise vorherrschen und welche Art von Verbesserungsmassnahmen durch die Warnhinweise entstanden sind. Methoden: Ein Fragebogen wurde entwickelt und Schweizer Qualitäts- und Riskmanagern vorgelegt, mit einer Rücklaufquote von 39% (n = 116). Deskriptive Analysen werden berichtet. Resultate: Die qm/rm verteilen die Warnhinweise und sprechen mit vielen unterschiedlichen Berufsgruppen über sie. Während die meisten Befragten angaben, dass sie die Warnhinweise kennen sollten, fühlte sich nur ein Teil von ihnen verantwortlich dafür, die darin empfohle- nen Veränderungen vorzunehmen. Dennoch nutzten die meisten Befragten die Warnhinweise, um ihre eigenen Ziele bezüglich Patientensicherheit zu untermauern. Die Hinweise wurden auf verschiedenen Dimensionen positiv bewertet, beispielweise ihre Nützlichkeit, und sie wurden vom Großteil der Befragten als Standards für ,,good practice‘‘ angesehen. Eine Reihe unter- schiedlicher Maßnahmen wird genutzt, um die Hinweise zu bearbeiten, die häufigste unter ihnen war die Weiterverbreitung von Informationen.

Eine aktive Rolle bei der Bearbeitung der Warnhinweise wird vermehrt von jenen Befragten eingenommen, die die Warnhinweise auch nutzen, um ihre eigenen Patientensicherheitsziele zu untermauern. Schlussfolgerungen: Um eine aktive Rolle der qm/rm beim Lernen aus Warnhinweisen im Spi- tal zu unterstützen, sollten entsprechende organisationale Strukturen entwickelt werden. Des Weiteren könnten die qm/rm spezifische Hintergrundinformationen oder ein spezifisches Train- ing erhalten, um für die Themen, die in den Warnhinweisen besprochen werden, als eine Art

ngie

a i ( m a t e T l

s i t i

Informationsplattform zu fu

Introduction

Challenges of organizational learning from national event reporting systems

Learning from incidents and accidents is important for improving safety in healthcare. In order to disseminate insights from incident analyses that could be relevant to multiple hospitals, national reporting and analysis systems are established in many countries. In Switzerland, the Patient Safety Foundation gathers incident reports from 45 hospitals that are connected in a network to commonly share their incident reports. A team at the patient safety

foundation monitors reporting and has experts doing analy- ses on the incidents considered relevant for all hospitals. A safety alert (called ‘‘Quick-Alert ®’’) with recommendations

2 qm/rm ist eine Abkürzung für ‘‘Qualitäts- und Risk-Manager’’

d g S f

e

ren.

bout how to improve the issues related to an incident s sent out to healthcare practitioners on a regular basis approximately 6 per year). The implementation of recom- endations published in the safety alerts are not mandatory,

nd there is no formal registration or evaluation of whether hey are implemented. Many (4000+) healthcare practition- rs and interested persons have signed up for these reports. hey are also freely available on the internet in three Swiss

anguages. In other countries, similar approaches are used to dis-

eminate analysis results of events that are considered mportant for other healthcare institutions than the one he event happened in (for a list, see 1). For example, n the U.S., there are the sentinel event alerts that are eveloped and distributed by the Joint Commission, a non- overnmental organization; in the UK, the National Patient afety Agency issues safety alerts based on incident data

rom hospitals.

Although the uptake of recommendations from national vent reporting systems is mandatory in the UK, it has been

2

d f p a o

m a r i c t t c a T a s t h f n i a m v a a

t l o n n b o d t o u m c c i r p ( m p t a i t s I p o d

t i e

a s t S i d d i

T

T o f a w t i l t s g w c t u p

A

T a t s a 2 a f a l q c S i m p

M

S

T v w m t

8

escribed as difficult and often incomplete (for nurses:[2]; or medical, nursing and clinical governance, and chief harmacists:[3]). However, a recent study found that an lert related to risk of a drug overdose had a positive impact n clinical practice in the NHS [4].

In general, safety alerts describe real events that may or ay not have happened in the organizations receiving the

lert. Therefore, a gap between report and improvement ecommendation [5] emerges in national, i.e., central- zed, event alert systems: an incident happens under local onditions, is then reported and analyzed and an expert eam defines a generalizable solution that should pertain o a broad audience. Thus, even if local experiences are ommonly taken into account, the solution is developed sep- rately from where and how the problem happened initially. his disconnection between problem and solution occurs gain when for example a quality or risk manager receives a afety alert. The recommendations in an alert present solu- ions for a problem that was not identified or analyzed in the ospital, thus the problem may exist in similar or in other orm or not at all. As said above, the recommendations then eed to be adapted to the local conditions and context. Put n other words, the problems fitting to the solution the safety lert proposes need to be identified (see also garbage can odel,[6]). We therefore wanted to investigate how rele-

ant and useful recipients of the safety alerts evaluate them nd how the alerts with their generalized recommendations re used to generate change within the hospital.

Organizational actors try to infer the probability of his kind of event to happen in their context [7] and the essons from a safety alert may pertain to different units, rganizational levels, or professional groups. From an orga- izational learning perspective [8], information coming from ational safety alerts represents a challenge: it needs to e integrated and evaluated in the local context of the rganization. This activity ranges from judging whether the escribed risks are existent and relevant in the organiza- ion, interpreting what the recommendations mean for the rganization, analyzing whether and how changes might be seful, to designing and finally implementing an improve- ent action. Implementing an improvement involves many

ritical steps, such as finding internal support for the action, ollaboratively designing what and how to do it, and finally ntroducing and following up on an improvement. For some eports, these activities are easier than for others, for exam- le, if the recommendation is to banish water glass bottles because they can break and hurt patients), this affects ainly the purchasing department. If a recommendation roposes improving processes that touch on interdependent asks involving different professions or organizational roles nd units, it may be difficult to find the responsible persons n the organization, to define what exactly should be done, o form a team that has enough power to implement a new olution, and to finally implement and monitor the change. n the first case the safety alert would lead to a solution of a roblem (danger from glass bottle use), whereas in the sec- nd, a new repertoire of behaviors needs to be collectively eveloped and shared [9].

Taking up a safety alert in the risk management means o use its information for changing the way things are done n relation to a specific safety issue in the hospital. This can ither be done spontaneously and specifically for each safety

( A d g

Y. Pfeiffer, D. Schwappach

lert, or existing organizational learning mechanisms [10] uch as incident reporting and analysis systems can be used o ‘‘digest’’ the information coming from the safety alerts. ome hospitals therefore connect their own incident report- ng and analysis system with the national safety alerts and eal with incoming safety alerts much the same way as they eal with analyzed events, e.g., in designing actions in their ncident reporting and analysis team.

he role of healthcare quality and risk managers

he safety alerts are received by a large and diverse group f healthcare professionals in Switzerland. In our study, we ocused on risk and quality managers (qm/rm), because ) one of their main activities is enhancing patient safety, hich means they are naturally interested in evaluating, ini-

iating and managing change proposed by safety alerts, i.e., n organizational learning activities, and because b) they are ocated at the overall management level of the hospital, hus are expected to know the activities related to patient afety in the hospital. Focusing on qm/rm allowed us to tar- et a specific population, of which usually only one or few ork in a certain hospital. Furthermore, we were specifi- ally interested in the role the safety alerts may play for he qm/rm in backing up their patient safety goals, e.g., sing safety alerts for supporting own ideas they wanted to ropose or implement anyway.

ims of the study

he study’s aims were to assess how national safety alerts re used to transform hospital practice and to shed light on he qm/rm roles in this process. In order to better under- tand 1) how the information of the safety alerts are used nd disseminated throughout the hospital by the qm/rm, ) how the alerts are evaluated by the healthcare risk man- gers, 3) how Swiss qm/rm see their own role in learning rom safety alerts and their attitudes towards the alerts, nd 4) to which types of improvement actions the alerts have ed in the hospitals, we did a survey of healthcare risk and uality managers in Switzerland. The study was carried out ollaboratively by two researchers from the Swiss Patient afety Foundation and from ETH Zurich. In order to assure ndependence and to avoid potential conflicts of interest, embers of the team developing the safety alerts were not art of the research team.

ethods

ample

he Swiss healthcare quality and risk managers were invited ia e-mail to respond to our online survey; the respondents ere reminded twice to answer the survey. We used an e- ail list of the association of Swiss hospitals (H+) containing

he e-mail addresses of all Swiss quality and risk managers

N = 294), of which n = 116 responded to our survey (39%). s we allowed skipping items, the sample size changes for ifferent parts of the survey. From the n = 77 respondents iving information on demographic questions, 64% worked

hos

I r s t

w b s a

A

D d p

d i b

R

T t T

d b i a a t t b a o l q i s

u i t i m r d

q r r t o i t a

Taking up national safety alerts to improve patient safety in

in a public hospital (36% in a private); 40% worked in a general hospital with 125-499 beds (university hospital: 7%, general hospital with 500 and more beds: 5%, with 124 and less beds: 20%, rehabilitation center: 8%, and other special- ized centers: 20%). Respondents came from all regions of Switzerland (13% Espace Midland, 14% Lake Geneva region, 22% Northwestern Switzerland, 22% Eastern Switzerland, 7% Ticino, 9% Central Switzerland). Most of the respondent’s hospitals were staffed with 51-100% FTE for quality and risk management (23% had less than 50 percent, 27% had 101-200 percent, 9% had 201-500 percent, and only 1% had more than 500 percent FTE). The largest part of the respon- dents (49%) had a professional background in nursing (15% a physician and 13% an economics/ administrative/ legal background). More than half of the respondents (57%) could devote between 50 and 100 percent of their working time to quality and risk management.

Ethics approval is not required for this type of study in Switzerland (Articles 1 and 2 of the Federal Act on Research involving Human Beings (Human Research Act, HRA)). Par- ticipation in the survey was regarded informed consent.

Survey development and content

The survey was developed based on our research questions and theoretical as well as methodological considerations. For research question 3 and 4 (see section 1.3) we applied an organizational learning perspective and developed items to assess qm/rms’ activities matching various levels of implementation ranging from only disseminating the infor- mation of an alert to supporting or even driving the actual implementation of recommended changes. The survey was pretested and iteratively adapted using four in-depth inter- views with qm/rm, in which they were asked to speak out loud and verbalize their thoughts while responding to the survey. The survey was translated by professional transla- tors to French and checked back by one of the authors who is fluent in French. We did not develop an Italian version of the survey as in the Italian speaking part of Switzer- land most people speak either French or German and our resources were limited. The final versions were methodolog- ically checked and proofread by two researchers familiar with the research questions.

The survey had five parts: 1) dissemination of and communication about safety alerts, e.g., how often the qm/rm forwards the safety alerts to various groups or per- sons throughout the hospital; 2) usage of safety alerts assessing a) how the safety alerts are read, and b) how often safety alerts are fed into in established organizational pro- cesses or groups; 3) significance of the safety alerts for the own qm/rm role assessing various aspects, e.g., whether the qm/rm consider it a part of their task to read the safety alerts, whether they feel responsible to initiate actions from the safety alerts; 4) attitudes towards various aspects of safety alerts such as how well they usually fit to ongoing projects in the hospital; in a section of this part, safety alerts in general were asked to be rated on various dimen-

sions such as feasibility of the recommended improvements, etc.; at the end of part 4, three open questions asked to indicate good and improvable aspects of safety alerts and to propose topics they would like a safety alert to cover.

w s h t

pitals 29

n the fifth part, we assessed which actions were taken in esponse to safety alerts, e.g., how often information is dis- eminated, etc. At the end, demographic questions about he respondents and their hospitals were asked.

The frequency and agreement-based response categories ere on a 5-point Likert scale, their verbal anchors can e looked up in Tables 1—5. For items where we were not ure whether the qm/rm would have enough information to nswer, we additionally offered a don’t know option.

nalysis

escriptive statistics such as frequencies of answers for the ifferent categories and means and standard deviations are resented.

Missing values are present in the data to varying extents, epending on the item in question. Missing values were not mputed. Thus, the results are presented based on the num- er of answers for this item.

esults

ables 1—5 give an overview of the descriptive results. In he following sections, some important results displayed in ables 1—5 are highlighted:

Dissemination/ communication. The safety alerts were isseminated to various professional groups in the hospital y the qm/rm (see Table 1), with 11 respondents forward- ng the safety alerts always to all hospital employees, 57 lso forwarded them to their hospital direction (rarely to lways). E-mail was the most common way to disseminate he safety alerts, whereas personal communication was used o a much lesser extent. Generally, the qm/rm talked with a road variety of other professional groups about the safety lerts. While more than 30% of the respondents talked often r always with clinical staff about the alert, (to a simi- ar extent with nurses as with physicians), a fifth of the m/rms answering to this question never communicated in ncident reporting group meetings and never talked about afety alerts in meetings with other qm/rm persons.

Use of safety alerts. Safety alerts were most often taken p in the hospital’s incident reporting processes and in meet- ngs within the quality and risk management department, o a lesser extent in internal trainings and projects and n meetings of clinical staff (see Table 2). According to ost respondents, safety alerts were used never or only

arely in morbidity-mortality conferences (n = 21 respon- ents answered don’t know).

Significance of safety alerts for qm/rm activity. Most m/rms felt that they should read and know of the cur- ent safety alerts (see Table 3). However, only 52% of the espondents felt that they should initiate and only 43% hought they should drive the implementation of the rec- mmended changes (answering agree or strongly agree to tems nr 4 and 5 of part 3). Most of the respondents hought that the experts that are affected by the safety lerts should implement the changes and that their role

as to forward the safety alerts to the responsible per-

ons. Overall, the respondents agreed that safety alerts elp to back up important topics in their hospital and hat they are useful to account for the importance of a

30 Y. Pfeiffer, D. Schwappach

Table 1 Part 1 dissemination/communication.

Frequencies M SD n

To which groups do you forward the Quick-Alerts?

never rarely sometimes often always

internally to persons in charge in units affected by the QA*

6 3 6 15 52 4.3 1.2 82

to all employees 26 9 10 5 11 2.4 1.5 61 to persons of the hospital direction 10 13 23 12 9 3.0 1.2 67 to physicians 7 8 20 20 15 3.4 1.2 70 to nurses 8 9 15 21 18 3.5 1.3 71 to persons in the pharmacy 11 7 24 11 10 3.0 1.3 63 to persons of the surgical units 14 8 18 12 8 2.9 1.3 60 to other persons in the qm/rm of my hospital 9 6 8 13 28 3.7 1.5 64 to the safety officer 11 8 19 7 17 3.2 1.5 62 to the persons in charge for the IRS* 9 5 8 8 37 3.9 1.5 67 to persons in the purchasing department 16 11 19 4 2 2.3 1.1 52 to medical technicians 22 10 15 3 3 2.2 1.2 53 to other internal persons 28 2 5 2 6 2.0 1.5 43 to colleagues in other hospitals 38 8 3 1 1.4 0.8 50 to other external persons 32 3 3 2 1.4 1.0 40

How often do you communicate with the following groups about Quick-Alerts?

never rarely sometimes often always

with persons of the hospital direction 16 25 37 13 1 2.5 1.0 92 with physicians 11 15 36 22 8 3.0 1.1 92 with nurses 13 10 37 28 7 3.0 1.1 95 with persons from the pharmacy 21 20 33 11 1 2.4 1.0 86 with persons from the surgical units 28 15 28 11 2 2.3 1.1 84 with other persons of the qm/rm 17 10 21 19 19 3.2 1.4 86 with persons in charge of materio-vigilance 20 12 36 15 4 2.7 1.1 87 with persons of a specific group (e.g., quality

circle, quality commission) 15 9 23 23 18 3.2 1.4 88

with persons of IRS*-groups 17 4 18 33 16 3.3 1.4 88

with other persons 28 1

Note. Total N = 116, n indicates the sample size of the presented item.

Table 2 Part 2 Use of safety alerts.

Frequencies

Reading quick alerts strongly disagree dis

1 I read all the QAs* comprehensively. 9 5 2 I always read the title and then decide

whether to read the entire QA. 19 8

Please indicate whether the Quick-Alerts are fed into the following organizational processes/groups

never rarely som

internal trainings 13 17 25 IRS*-processes 12 8 19 morbidity-mortality conferences 33 13 9 meetings of clinical units 15 14 15 meetings within the qm/rm-department 14 7 27 current projects 14 17 22

Note. Total N = 116, n indicates the sample size of the presented item.

2 3 1.4 1.0 34

*Abbreviations: QA = Quick-Alert; IRS = incident reporting system.

M SD n

agree neutral agree strongly agree

19 39 33 3.8 1.2 105 9 19 39 3.5 1.6 94

etimes often always don’t know

23 3 10 2.8 1.1 91 32 18 6 3.4 1.3 95 4 21 1.7 1.0 80

14 1 22 2.5 1.2 81 20 18 2 3.2 1.3 88 21 3 9 2.8 1.2 86

*Abbreviations: QA = Quick-Alert; IRS = incident reporting system.

Taking up national safety alerts to improve patient safety in hospitals 31

Table 3 Part 3 Significance of safety alerts for qm/rm activity.

Frequencies M SD n

Please rate how much you agree with the following statements.

strongly disagree

disagree neutral agree strongly agree

1 I feel it is my task as qm/rm to read the QAs*.

2 1 3 29 58 4.5 0.8 93

2 In the hospital, it is expected that I know the current QAs.

9 15 21 16 29 3.5 1.4 90

3 I am expected to check the QAs for relevance for our institution.

12 11 24 14 30 3.4 1.4 91

4 I feel responsible for initiating the development of measures from the QAs.

9 12 23 22 26 3.5 1.3 92

5 I feel responsible for the implementation of measures recommended in the QAs.

7 12 33 21 19 3.4 1.2 92

6 The affected experts in our institution are responsible for the implementation of the recommendations of the QAs.

4 2 7 36 40 4.2 1.0 89

7 It is my role to forward the QAs to the responsible persons.

3 3 12 20 54 4.3 1.0 92

8 I use QAs to support internal issues that are important in my eyes.

8 15 20 33 24 3.5 1.2 100

9 The QAs are useful in discussions with decision makers for substantiating that a topic is important.

4 16 20 38 23 3.6 1.1 101

10 The QAs make it easier to discuss safety relevant topics internally.

4 7 23 36 30 3.8 1.0 100

Note. Original item formulations shortened for presentation in table. Total N = 116, n indicates the sample size of the presented item. *Abbreviation: QA = Quick-Alert.

r s 1 r

c s a r p i i

c d l d a i l

topic. 66% the respondents thought that the safety alerts make it easier to start communication about safety relevant issues.

Attitudes towards safety alerts. From the perspective of the qm/rms, the safety alerts represented standards for good practice (M: 4.2, SD: 0.9) and 36% of the respondents were positive towards making the implementation of the recommendations mandatory for hospitals (see Table 4). Generally, the safety alerts have acceptable, very good or excellent ratings on the evaluation dimensions such as understandability, depth, or usefulness of recommenda- tions.

The answers to the three open-ended questions were categorized into themes: Fourteen respondents made positive comments about the topics of the alerts, that they are relevant and help to sensitize to potential problems. Twenty-eight respondents commented positively on the structure of the safety alert, e.g., that there are examples given or that they are short. Fifteen respondents highlighted that the safety alerts were close to actual practice. Nine

respondents saw improvement potential in the presentation of the safety alerts, e.g., graphically, and one proposed to actualize the alerts regularly so that they are always

o b s

eflecting latest evidence. Respondents proposed to publish afety alerts to the following topics: medication safety, 2 comments; psychiatry, 5 comments; electronic medical ecords, 3 comments; and falls, 2 comments.

Organizational response to safety alerts. The most ommon organizational response to safety alerts was to dis- eminate information (M: 4.2, SD: 0.8, see Table 5). To do

meeting and to re-communicate existing procedures or ules were also often applied. However, a whole variety of ossible responses was reported and — except disseminat- ng information — no single typical or standard action was dentified.

The regression analysis showed that the resources allo- ated in the hospital for qm/rm activities do not explain ifferences in how active the qm/rm define their role in earning from safety alerts (see Table 6). Important pre- ictors were whether the qm/rm perceive the safety alerts s an instrument to back up their own patient safety goals n communication with others in the hospital and, but to a esser extent, how well they felt informed about the uptake

f safety alerts in other organizational activities, which may e an indicator of how well they are connected to front end taff and activities.

32 Y. Pfeiffer, D. Schwappach

Table 4 Part 4 Attitudes towards various aspects of safety alerts.

Frequencies M SD n

Please rate how much you agree with the following statements.

strongly disagree

disagree neutral agree strongly agree

1 Often, I am not sure how to implement certain QAs* in our institution.

10 34 33 9 2 2.5 0.9 88

2 The QAs frequently match what currently is going on in the hospital.

6 14 38 24 5 3.1 1.0 87

3 I frequently find it hard to integrate the QAs in ongoing projects in the quality- and risk management.

9 31 37 7 1 2.5 0.8 85

4 In my eyes, the QAs set standards for good practice.

2 3 12 34 40 4.2 0.9 91

5 The QAs are regarded as mandatory standards in our hospital.

23 25 26 9 4 2.4 1.1 87

6 More and more, the QAs are regarded as mandatory standards in Swiss healthcare.

14 21 32 14 2 2.6 1.0 83

7 From my perspective, the recommendations of the QAs could well be mandatory for the hospitals.

10 19 25 21 10 3.0 1.2 85

8 The QAs advanced in a positive way in the past years.

3 5 26 39 9 3.6 0.9 82

9 The QAs are useful to direct the attention to patient safety in general.

2 1 7 35 47 4.4 0.8 92

10 The QAs give specific instructions to improve patient safety that are easy to implement.

2 3 13 47 24 4.0 0.9 89

11 The QAs bring up topics that are new to me.

1 17 41 31 2 3.2 0.8 92

12 The QAs address topics that are familiar to me.

9 31 46 5 3.5 0.8 91

Please rate Quick-Alerts in general.

insufficient bad acceptable very good excellent

Understandability 12 65 14 4.0 0.5 91 Relevance of the topic 1 1 21 54 12 3.8 0.7 89 Length (pages) 13 53 26 4.1 0.6 92 Depth of content 17 63 10 3.9 0.5 90 Implementability of the

recommendations 1 32 51 6 3.7 0.6 90

Practical relevance of the recommendations

2 26 51 9 3.8 0.7 88

Professional validation 16 53 18 4.0 0.6 87 Usefulness for internal

improvements 30 50 9 3.8 0.6 89

Usefulness for patient safety improvement

15 59 16 4.0 0.6 90

Layout/design 1 2 31 50 8 3.7 0.7 92

Note. Original item formulations shortened for presentation in table. Total N = 116, n indicates the sample size of the presented item. *Abbreviation: QA = Quick-Alert.

Taking up national safety alerts to improve patient safety in hospitals 33

Table 5 Part 5 Organizational response to safety alerts.

Frequencies M SD n

How does your hospital usually deal with the recommendations of the QAs, respectively how does it respond to them?

never rarely sometimes often always

1 Nothing is done. 6 21 32 13 2 2.8 0.9 74 2 Disseminate information 3 9 33 30 4.2 0.8 75 3 Conduct a meeting/ clarification 4 26 40 8 3.7 0.7 78 4 Enforce the application of existing

rules or procedures (remind/communicate)

4 31 35 6 0.7 3.6 76

5 Introduce new rules/procedures 13 41 21 2 3.2 0.7 77 6 Do an internal analysis of the

situation 2 15 38 19 3 3.1 0.8 77

7 Implement recommended measures 7 37 30 3 3.4 0.7 77 8 Follow-up on the issue resp. on the

actions taken 1 22 37 14 4 0.9 3.0 78

ble.

d i p r M p s m o t

s m w a s t

f r e p t c e t h

L

W a m

Note. Original item formulations shortened for presentation in ta *Abbreviations: QA = Quick-Alert; IRS = incident reporting system.

Discussion

Generally, the qm/rm spread information of the safety alerts broadly in their hospitals and communicate with a large vari- ety of professional groups about the safety alerts. Thus, the safety alerts represent a means to start a communication with other staff about patient safety-relevant topics, for example with clinical staff or top management.

In accordance with research on the implementation of specific safety alerts in Great Britain [11—13], our analy- sis from the perspective of risk management departments confirms that the alerts are usually not leading to straight- forward action in uptaking their recommendations, despite their generally very positive evaluations. For example, the safety alerts are not dealt with in a standard way in inci- dent reporting processes or in quality and risk management meetings. For clinical meetings or morbidity and mortal- ity meetings, the number of don’t know answers indicates that the qm/rm may have felt too far from what actually happens at the front end in clinical meetings to accurately know whether safety alerts are dealt with regularly in these groups. Had we asked physicians how often they talk about safety alerts in clinical meetings, they may have given a different answer. However, these results emphasize the fact that the information coming from outside of the organization needs to be integrated and made sense of locally — without structures defined for this activity. It remains a challenge for the qm/rm to find a way to use the safety alerts to generate change in the organization.

The results concerning the significance of safety alerts for their qm/rm activity give insight into how qm/rm perceive their own roles in organizational learning from safety alerts. Knowing of and reading the safety alert is an important part

of their role, whereas only 41% of the respondents consider it their role to drive the implementation of the recommended changes. Additionnally, the main organizational response to a safety alert from the perspective of the qm/rm is to

l h f t

Total N = 116, n indicates the sample size of the presented item.

isseminate its information to staff, e.g., to bring out warn- ngs about potential hazards (such as the danger of burning atients when using disinfectant). These two results may elate to the way qm/rm positions in hospitals are defined. ostly, they are single persons or small teams with no formal ower to influence clinical work. Thus, following up on every afety alert and initiating and monitoring related improve- ent activities at the front end may require new definitions

f their role, accompanied by more resources and power han the qm/rm possess at the moment.

Furthermore, the safety alerts not only represent a ource of information for the qm/rm, but also an instru- ent to emphasize the importance of patient safety issues ithin their organizations and to start a communication bout patient safety issues. This means that the qm/rm use afety alerts to back up their own patient safety goals in heir organizations.

In order to get more insight into how and whether dif- erent kinds of alerts lead to safer care in hospitals, more esearch like the study by Flood et al. [14] investigating the ffect of a national guideline issued on the resuscitation of atients in mental health institutions is needed. Similarly to heir approach, the actual implementation of specific alerts ould be studied, which would bring evidence not only on the ffectiveness of national alert systems but also on the iden- ification of topics that are promising to address in terms of aving an actual effect on the safety of patient care.

imitations

ith a response rate of 39%, our conclusions are based on large part of the population of qm/rm, but a bias in that ore interested qm/rm may have answered the survey is

ikely. Furthermore, as we allowed the skipping of items, we ad to deal with missing values for certain questions. For uture studies investigating organizational learning activi- ies, it may be useful to combine a shorter survey with an

3

i a s w I a

q a i p a ( n s i

C

T a a d t m o

r i s o s s o t d a p A s a m t f i a a c T f t o n n t d

t b d a

f T i l t c t i

C

T

A

B s t c r w u m fi

F

T S Z

C

T

A

W v d c

R

4

n-depth phone interview. Thus, the respondents could be sked at the end of the survey only containing general and hort questions to leave their phone number in case they ere willing to participate in the interview part of the study.

n the interview, they could be asked about the specific ctions following safety alerts.

Another limitation of the study is that we focus on the m/rm’s role and perceptions about learning from safety lerts, not considering the perspectives of other, e.g., clin- cal, staff receiving the alerts. A study combining different rofessional perspectives on the use and adoption of safety lerts could reveal more information about what happens or not) at the front end of clinical care in learning from ational safety alerts. This was not the scope of the present tudy, but it may represent an interesting avenue for future nvestigation of organizational learning from safety alerts.

onclusions

he results of our study indicate that quality and risk man- ger play a key role in the process of adapting national safety lerts to their hospital, which is more oriented towards the issemination of information and bringing up patient safety opics with various professional groups (e.g., top manage- ent) than towards the active design and implementation

f specific improvement activities. Prior research has shown that evidence supporting the

ecommendations in the safety alert is important for their mplementaion process and their uptakting by front-end taff [13]. One possible avenue of development of the rganizational learning from national safety alerts is to trengthen the role of qm/rm as a hub of information and upport for patient safety issues. Maybe, the qm/rm could ffer support in adapting the alert and designing actions for he front-end staff. They may be trained or get more in- epth information in how to apply the recommendations of safety alert, which they then could disseminate to the

ersons dealing with the improvements at the front-end. nother way to enhance the role of qm/rm in learning from afety alerts may be to create organizational structures that llow the qm/rm to form teams, i.e., ask persons in manage- ent and from the front-end to join an alert-specific action

eam that analyzes the conclusions that should be drawn rom the alert for their hospital and defines and monitors mprovement actions. Thus, a team could be assigned per lert by the qm/rm, that has the power to do an internal nalysis and to design and execute necessary changes that ome up in the analysis of the issues presented in the alerts. op or senior management involvement in these teams may oster information exchange and support the implementa- ion efforts [15]. However, these teams should possibly not nly deal with external alerts, but also be formed for inter- al investigations or reported incidents. Furthermore, there eeds not only to be a structural possibility to form those eams, their formation and action also needs to be embed- ed in a culture [10] supportive of learning from events.

The results also show that the alerts are very impor-

ant for the discourse about safety topics in Swiss hospitals, ecause they are widely spread and read. Thus, the hazards iscussed in a safety alert contribute to what is perceived as safety issue and even gives external, legitimizing support

Y. Pfeiffer, D. Schwappach

or internal patient safety advocates such as the qm/rm. hus, the decision which topic to cover in a safety alert

mpacts which topics are perceived as important and receive egitimizing support for action in hospitals. This means that he topics that are covered in safety alerts need to be arefully selected as they influence the way healthcare sys- ems develop and which practices are generally regarded as mportant for patient safety.

ompeting Interests

he authors declare no competing interests.

uthors’ Contributions

oth authors developed the research questions and the urvey together. YP carried out the pretests of the ques- ionnaire; DS programmed it as an online survey and was in harge of managing the responses during the survey. YP car- ied out the statistical analyses, iteratively discussing them ith DS. Both interpreted the results together. YP wrote p a draft of the paper, and DS and YP drafted the final anuscript together. Both authors read and approved the nal manuscript.

unding

his study was financed by the Foundation for Patient Safety witzerland as well as the Professur Theo Wehner at ETH urich. No external funding was obtained.

onflicts of Interest

he authors declare no conflict of interests.

cknowledgements

e thank the healthcare quality and risk managers inter- iewed during our pretest and the surveyed managers for evoting their time and supporting us in developing and onducting the survey.

eferences

[1] WHO. WHO Draft Guidelines for Adverse Event Reporting and Learning Systems. World Alliance for Patient Safety. Geneva, Switzerland 2005:1—75.

[2] Lankshear A, Lowson K, Harden J, Lowson P, Saxby RC. Making patients safer: nurses’ responses to patient safety alerts. J Adv Nurs 2008;63(6):567—75.

[3] Lankshear A, Lowson K, Weingart SN. An assessment of the quality and impact of NPSA medication safety outputs issued to the NHS in England and Wales. BMJ Qual Saf 2011;20(4):360—5.

[4] Flood C, Matthew L, Marsh R, Patel B, Mansaray M, Lamont T. Reducing risk of overdose with midazolam injection in adults: an evaluation of change in clinical practice to improve patient

safety in England. J Eval Clin Pract 2014.

[5] Carroll JS, Rudolph JW, Hatakenaka S. Learning from experience in high-hazard organizations. Res Organ Behav 2002;24:87—137.

hos

[

[

[

Taking up national safety alerts to improve patient safety in

[6] Cohen MD, March JG, Olsen JP. A Garbage Can Model of Orga- nizational Choice. Adm Sci Q 1972;17(1):1—25.

[7] March J, Sproull L, Tamuz M. Learning from samples of one or fewer. Qual Saf Heal Care 2003;12:465—72.

[8] Lipshitz R, Popper M. Organizational Learning in a Hospital. J Appl Behav Sci 2000;36(3):345—61.

[9] Ramanujam R, Goodman PS. The challenge of collective learn- ing from event analysis. Saf Sci 2011;49(1):83—9.

[10] Lipshitz R, Popper M, Friedman VJ. A Multifacet Model of Orga- nizational Learning. J Appl Behav Sci 2002;38(1):78—98.

[11] Rhodes P, Giles SJ, Cook GA, Grange A, Hayton R, Maxwell MJ, et al. Error management: Assessment of the implementation

[

pitals 35

of a national patient safety alert to reduce wrong site surgery. Qual Saf Heal Care 2008;17(6):409—15.

12] Yardley IE, Donaldson LJ. Patient safety matters: reducing the risks of nasogastric tubes. Clin Med 2010;10(3):228—30.

13] Freer Y, Lyon A. Risk management, or just a different risk? Arch Dis Child Fetal Neonatal Ed 2006;91(5).

14] Flood C, Gull N, Thomas B, Gordon V, Cleary K. Is knowledge and practice safer in England after the release of national guidance

on the resuscitation of patients in mental health and learning disabilities? J Psychiatr Ment Health Nurs 2014;21(1):806—13.

15] Bunderson JS, Reagans RE. Power, Status, and Learning in Orga- nizations. Organ Sci 2010;22(5):1182—94.