WRITTEN REPORT Study Programme in Medicine
Master’s thesis (30c)
What physicians, managers and other professionals involved in Swedish National Quality Registries consider as barriers and facilitators to
improve patient care with help of registries and their output
By: Sanna Björk Runsvik
Supervisor: Ann Catrine Eldh, RN, PhD, Associate Professor
Date: 2016-03-24
Title: WHAT PHYSICIANS, MANAGERS AND OTHER PROFFESSIONALS
INVOLVED WITH SWEDISH NATIONAL QUALITY REGISTRIES CONSIDER AS BARRIERS AND FACILITATORS TO IMPROVE PATIENT CARE WITH HELP FROM REGISTRIES AND THEIR OUTPUT.
By: Sanna Björk Runsvik
Contents
ABSTRACT ...3
POPULÄRVETENSKAPLIG SAMMANFATTNING [IN SWEDISH] ...4
BACKGROUND ...5
METHODS ...10
RESEARCH DESIGN ...10
SAMPLE AND PROCEDURE ...10
DATA COLLECTION AND MEASURES ...11
DATA ANALYSIS ...14
RESULTS ... 15
DISCUSSION ... 21
VALUE OF REGISTRIES FOR HEALTH CARE ... 22
DIFFERENCES IN PERCEPTION OF QUALITY REGISTRIES BETWEEN PROFESSIONALS ... 23
VALUE OF RESOURCES FOR NQRS ... 24
AIMS AND RESULTS ... 24
METHODOLOGICAL CONSIDERATIONS ... 24
CONCLUSION ... 26
ACKNOWLEDGMENTS ... 26
REFERENCES ... 27
ATTACHMENTS ... 29
ATTACHMENT A.INTERVIEW GUIDE [IN SWEDISH]... 29
ATTACHMENT B.CATEGORIES FORMED BY THE ANALYSIS OF THE QUESTIONNAIRE DATA ... 30
Abstract
Sweden has 96 National Quality Registries (NQRs), which are supposed to contribute to clinical quality improvement. However, if and to what extent they benefit health care quality is not known.
The purpose of this study was to increase the knowledge of the relationship between quality improvement and NQRs, seeking to elucidate obstacles and facilitators for quality improvement related to context and NQRs. Further, it investigated differences between the experience of physicians and other professionals.
127 individuals provided free text comments to a validated national survey on three NQRs, and six additional interviews were conducted with key informants. All texts were analyzed using content analysis.
The findings show that the NQRs are to some extent utilized for quality improvement work. Yet, valid data, resources to analyze outcomes, and resources for quality improvement are scarce.
While staff engaged in the NQRs considers them beneficial, physicians have the most use of them.
However, registry structures and incentives need to improve, and a wider engagement in the NQRs is required.
Registries support quality improvement but their potential is far from utilized. In particular, there is a need of allocated resources for data management and committed professionals on all levels of NQRs.
Populärvetenskaplig sammanfattning [in Swedish]
I Sverige finns idag 96 nationella kvalitetsregister, NKRs. Registren är strukturerade databaser med individbaserad information om hälsoproblem och deras åtgärder och resultat. Kliniskt förbättringsarbete är ett av huvudändamålen med NKR men om och hur detta sker är inte
fullständigt klarlagt. Snarare är relationen mellan kvalitetsregister och förbättringsarbete osäker.
Den här studien är en del av ett större forskningsprojekt av om och hur vårdpersonal involverade i NKRs använder och uppfattar NKRs samt registrens bidrag till förbättringsarbete. En
enkätundersökning är gjord bland kliniska användare av tre NKRs. Förutom skattningar fanns möjlighet till fritextkommentarer till frågorna, vilka har analyserats i denna studie. För en fördjupad förståelse genomfördes också intervjuer med nyckelpersoner involverade i de tre registren. Deltagarna fick reflektera över vårdens kvalitet, registerdata, organisatoriska
förutsättningar, egen och verksamhetens användning av NKR och registrets betydelse. Texterna från enkät- och intervjusvar analyserades med innehållsanalys. Resultatet visar att datavaliditet var viktigt för att deltagarna skulle använda registret i förbättringsarbete. Andra viktiga faktorer var regelbunden analys och presentation av utfall, att integrera registrering i det kliniska
vardagsarbetet. Dock behövdes mer resurser till registerarbete samt en uppdaterad registerstruktur, gärna med koppling till journalsystem. Användarna var generellt positiva till register och de ansågs ha en viktig roll i förbättringsarbete även om användningen behöver förbättras.
Vårdpersonal och nyckelpersoner behöver avsatt tid för att underlätta den processen men också andra berörda, som kollegor, chefer och beslutsfattare, behöver bli mer engagerade i registren och nyttja dem för kvalitetsarbete. Resultaten sammanfaller med andra studiers resultat och
understryker vikten av att tillvarata den unika resurs som svenska NKRs utgör.
Background
There are around 96 national quality registries, NQRs, operating in Sweden that are financed by the public health care authorities and the state1. The following is a description of a national quality registry according to the Swedish municipalities and county councils: “A National Quality Registry contains individualized data concerning patient problems, medical interventions, and outcomes after treatment, within all healthcare production. It is annually monitored and approved for financial support by an Executive Committee.”2 As stated by The National Board of Health and Welfare, quality work improvement is one of the intended uses of a national quality registry.3 Further, the Swedish municipalities and county councils states that “the registries are constructed by the professional groups in that they themselves are to benefit from the NQRs in their everyday work.”4
Quality is a key word in the NQRs. A general definition of quality, separate from health care quality, from the International Organization for Standardization is: “[…] all the properties of a product that gives it its ability to satisfy expressed or implied needs […]”.5 The definition of quality is evidently wide and can be placed in different contexts. In terms of health care quality, the World Health Organization (WHO) began health care quality work in the beginning of 1980 and characterizes quality in health care and health systems as multi-dimensional. Six dimensions can be distinguished as requirements of how health care should be: effective, efficient, accessible, acceptable/patient-centered, equitable and safe. These are dimensions applicable to a general system perspective and thus relevant internationally.6 The Swedish National Encyclopedia defines health care quality as “the level of accomplished goals in health care works.” The Swedish
definition therefore also focuses on the goal/aim with health care. The goal of health care quality is described as “fulfilling the patients’ expectations and wishes as well as, as far as possible, not only alleviate symptoms and cure disease but also give good physical and mental care.”7 The Swedish National Board of Health and Welfare is less precise than the dictionary and refers to the
1Karin Christensson, “Nationella Kvalitetsregister - SKL,” Sveriges Kommuner Och Landsting, accessed January 27, 2016,
http://skl.se/halsasjukvard/kunskapsstodvardochbehandling/kvalitetsregisternationella.
2 Local Authorities and Regions SKL, “Quality Registries in Sweden,” In English - Kvalitetsregister, accessed January 27, 2016, http://www.kvalitetsregister.se/sekundarnavigering/inenglish.
3 Socialstyrelsen, “Nationella Kvalitetsregister,” Nationella Kvalitetsregister, accessed January 27, 2016, https://www.socialstyrelsen.se/register/registerservice/nationellakvalitetsregister.
4 Christensson, “Nationella Kvalitetsregister - SKL.”
5 Bengt Lundh, “Nationalencyklopedin, Vårdkvalitet.,” Nationalencyklopedin, accessed February 10, 2016, http://www.ne.se/uppslagsverk/encyklopedi/lång/vårdkvalitet.
6 R Bengoa et al., Quality of Care: A Process for Making Strategic Choices in Health Systems. (Geneva: World Health Organization, 2006), http://www.who.int/management/quality/assurance/QualityCare_B.Def.pdf.
7 Lundh, “Nationalencyklopedin, Vårdkvalitet.”
law, defining that “a health organization should achieve goals and requirements as stated by the law and other regulations and also those resolutions supported by them to achieve quality. “8
To achieve desirable health care quality appropriate to existing knowledge and experience quality improvement work is requested. Improvement work is having strategic plans and making informed choices to improve quality. WHO emphasizes two main arguments as to why efforts are needed in this field:
1. Well developed and well sustained health systems do not achieve the expected outcomes and the standard of health care varies both within and between health care systems making quality a concern
2. In health systems with fewer resources, particularly in developing countries, improvement processes need to be based on local strategies to achieve the best possible results,
optimizing the use of limited resources. 9
In Sweden, systematic management systems for health care quality improvement are regulated by The National Board of Health and Welfare’s regulation10. The regulation propose that “the intended use of such a management system is to systematically and continuously develop and secure the quality of the organization as well as to plan, lead, control, follow up, evaluate and improve the organization.”11
In general, quality improvement work is something that aims to adjust health care in a way that will give better patient outcomes, better system performance and better professional development.
The adjustments involve changes and require engagement from health care professionals amongst others. 12 Amory Codman, a surgeon at Harvard in the beginning of the 20th century, is known to have developed and implemented “The End Result Idea”; He proposed that the hospitals ought to have follow ups on patient treatment to learn the end result, which consecutively would be crucial for improvement. He was also the person to start the first cancer registry in the United States of America, in line with his beliefs of acquiring data/results for analysis, comparison and
subsequently to achieve improvements. The first Swedish quality registry started in 1975, on knee
8 The National Board of Health and Welfare in Sweden, Ledningssystem För Systematiskt Kvalitetsarbete, Socialstyrelsens Författningssamling, vol. SOSFS 2011:9, accessed February 22, 2016,
https://www.socialstyrelsen.se/Lists/Artikelkatalog/Attachments/18389/2011-6-38.pdf.
9 Bengoa et al., Quality of Care: A Process for Making Strategic Choices in Health Systems.
10 The National Board of Health and Welfare in Sweden, Ledningssystem För Systematiskt Kvalitetsarbete.
11 Ibid.
12 P. B Batalden and F. Davidoff, “What Is ‘Quality Improvement’ and How Can It Transform Healthcare?,” Quality and Safety in Health Care 16, no. 1 (February 1, 2007): 2–3, doi:10.1136/qshc.2006.022046.
arthroplasty. In 1993, representatives from The Swedish councils and The National Board of Health and Welfare formed a steering committee for National quality registers, agreeing on how to form and fund new registries, thus laying the foundations for the decentralized Swedish registry structure. 13
The introduction of the knee arthroplasty registry along with the subsequent hip arthroplasty registry, along with the national analysis and comparison of the clinics’ results, gave Sweden a world-leading position in terms of hip- and/or knee replacement health care. 14 It is one example of how data generated from NQRs can be used for different research projects 15 which naturally could support quality improvement in itself. However, in sheer research there are controlled variables whereas quality improvement is about change and actions based on experiences in complex clinical systems. Further, quality improvement is prospective and does not rely on a search for new knowledge but rather on the ambitious efforts of both individuals and teams to make changes to protocol and to develop and test ideas. The context is of immense importance given the vast environmental settings in health care when making changes. The structure of the team, the specific care given and such parameters have to be taken into consideration. More specifically the Royal College of Physicians in London distinguish three dimensions to quality improvement: setting standards, measuring for quality, and quality improvement. The first dimension involves “defining expectations, levels of practice or outcomes”, and the second is about evaluation of practice and outcomes. The third dimension, named quality improvement itself, is “an active process that involves taking action, making changes and measuring progress towards an agreed aim.” 16
How do the Swedish NQRs contribute to their intended use as an instrument for quality work improvement? Despite Sweden’s vast accumulation of NQRs, the step between a registry and making it work as an aspect of quality improvement work in a particular organization is not apparent. The correlation between an NQR and quality improvement work is uncertain, both in
13 Sveriges Kommuner och Landsting, “Kvalitetsregistrens historia,” Kvalitetsregistrens historia - Kvalitetsregister, September 7, 2015,
http://www.kvalitetsregister.se/sekundarnavigering/omnationellakvalitetsregister/kvalitetsregistrenshistoria.293.html.
14 Ibid.
15 Sveriges Kommuner och Landsting, “För dig som forskar på registerdata,” Forska på registerdata - Kvalitetsregister, October 26, 2015, http://www.kvalitetsregister.se/forskning/forskaparegisterdata.1391.html.
16 Stephen Atkinson et al., “Defining Quality and Quality Improvement,” Clinical Medicine 10, no. 6 (December 1, 2010): 537–39, doi:10.7861/clinmedicine.10-6-537.
terms of the mechanism and the extent to whether or not this takes place.17 Internationally the term quality registry is seldom used with exception of Australia and Netherlands. And while the use of registries is common, the names and purposes vary.18 One example is the term medical registry which has a purpose of systematically and continuously collecting data for patients with specific health characteristics. 19 Another example is clinical registries with the purpose of collecting data on individual patients, rather than aggregated group data. National clinical databases in England contribute to quality improvement and micro-level evaluation of health technologies but the use on a meso-level in the organization, or at a macro-level for national policy, as in Sweden, is
uncertain. 20 However, some studies have been made of the relationship between registries and care improvement. A small study in USA in 2006 assessed if processes or structural characteristics of stroke programs could be related to the use of reports from a prototype registry to improve care.
To identify any relationships between the registry and positive quality indicators, the hospitals had to have an interest in the registry, in this case the monthly reports. This seemed to be “[…] related to skills required to make positive changes happen in stroke care.” Secondly, there had to be a translation of ongoing findings into potential care improvements where the researchers identified three main characteristics for accomplishment. These were “[…] documentation of care processes across departments; access to local or remote stroke teams; and data collection experiences […]”.
Due to the small size of the study and a risk of bias, these findings are primarily indications.21 Further, the susceptibility of systems dealing with data, like registries, to general practitioners was studied by another research group in Australia in 2010. Clinical data extraction programs/tools were offered to the participants to facilitate management of data. The intervention was built on an assumption that apart from motivation, general practitioners would need skills to extract and review data. The tools proved useful but one major issue identified was the insufficient basic data on patients, like allergies or smoking habits, indicating yet another important issue (apart from use of data): the necessity of covering input rather than extraction. Apart from absent data there was
17 Ann Catrine Eldh et al., “Facilitators and Barriers to Applying a National Quality Registry for Quality Improvement in Stroke Care,” BMC Health Services Research 14 (2014): 354, doi:10.1186/1472-6963-14-354.
18 Sofie Vengberg et al., “Användning Av Kvalitetsregister För Kvalitetsförbättring,” In Manuscript, n.d., 36.
19 Sabine N. van der Veer et al., “Improving Quality of Care. A Systematic Review on How Medical Registries Provide Information Feedback to Health Care Providers,” International Journal of Medical Informatics 79, no. 5 (May 2010): 305–23, doi:10.1016/j.ijmedinf.2010.01.011.
20 Nick Black and Stefanie Tan, “Use of National Clinical Databases for Informing and for Evaluating Health Care Policies,” Health Policy 109, no. 2 (February 2013): 131–36, doi:10.1016/j.healthpol.2012.10.007.
21 Valerie T. Stewart, “Use of a Prototype Acute Stroke Registry to Improve Care: Profile of Receptive Stroke Programs,” American Journal of Preventive Medicine, Paul Coverdell National Acute Stroke Registry
PrototypesAssessing Acute Stroke Care in the U.S. and Beyond, 31, no. 6, Supplement 2 (December 2006): S217–23, doi:10.1016/j.amepre.2006.08.014.
also important that registered data was accurate and complete. 22 Rather, the validity of the data is a re-occurring concern in several studies23, both in Sweden and internationally, and in all
probability an obstacle for optimizing the use of quality registries for quality improvement.
While a number of factors are known as to what might limit the application of quality registries for quality improvement, the full picture is yet to come. For example, there is limited knowledge as to what is the intended use as a tool for quality improvement – is it an ambition or a reality in the everyday practice of health care today? This study intends to broaden the knowledge of the relationship between quality improvement and NQRs. The aim of the study is to identify what physicians, managers and other professionals involved with NQRs consider as obstacles and facilitators for quality improvement work related to NQRs. While the international aspects of accuracy of registry data and other things related to registries are highly relevant, there are also differences in terms of context and intended use24. Thus, this study focuses on registries in Sweden, more precisely three NQRs and the professionals working with them, investigating barriers and facilitators to NQRs and clinical health care improvement. In particular, the study focuses on the following research questions:
What do professionals describe as factors affecting quality improvement related to quality
registries? Particularly, what is their experience of: health care quality and resources; registry data;
organizational conditions for registry work, and; their own application and the organizations applications of registry data, respectively?
What differences and similarities are there between the factors that physicians describe as
obstacles and facilitators for quality improvement, compared to those described by managers and other professionals?
22 Peter Schattner et al., “Clinical Data Extraction and Feedback in General Practice: A Case Study from Australian Primary Care,” Journal of Innovation in Health Informatics 18, no. 3 (2010): 205–12.
23 E H Bradley, “Data Feedback Efforts in Quality Improvement: Lessons Learned from US Hospitals,” Quality and Safety in Health Care 13, no. 1 (February 1, 2004): 26–31, doi:10.1136/qhc.13.1.26; Schattner et al., “Clinical Data Extraction and Feedback in General Practice”; Eldh et al., “Facilitators and Barriers to Applying a National Quality Registry for Quality Improvement in Stroke Care.”
24 Vengberg et al., “Användning Av Kvalitetsregister För Kvalitetsförbättring.”
Methods
Research design
This was a qualitative research study assessing obstacles and possibilities for quality improvement work. This study is part of a larger research project by the Department of Public health and Caring sciences, University of Uppsala, by orders of the Swedish municipalities and county councils. This part of the project focused on free text comments from a questionnaire distributed to professionals involved in three national quality registries in Sweden. In addition to the questionnaire comments, semi-structured interviews were conducted with professionals involved in the same three registries in one of the participating hospitals. The questionnaire assessed elements enhancing
improvements of quality in health care relating to quality registries regarding: health care quality and resources, data from the quality registry, organizational conditions for registry work and their own use or the organization’s use of data from the registry. The interviews focused on the same issues, but provided a more thorough understanding of the professionals’ experiences.
Sample and Procedure Questionnaire
The questionnaire was distributed to healthcare professionals working with the three Swedish national quality registries, NQRs. The registries included were the national registry on stroke, Riksstroke, the registry for gall bladder surgery, GallRiks, and the lung cancer registry, Nationellt lungcancerregister. Every health care unit working with these three NQRs received a
questionnaire. Respondents were identified with support from each registry’s administration center and included professions as: the head of the clinic, the physician(s) in charge of the local registry work, nurses and medical secretaries (if any) involved in the local registry work. Each person was contacted via email. In the cases of no response, corresponding reminders were sent out up to three times with a fourth final reminder including an opportunity to bestow motives for not
partaking. Voluntary completion and submission of the survey was considered individual consent.
A detailed disclosure of how respondents were identified and contacted is available.25
Interviews
For the interviews, key informants, i.e. the corresponding professionals working with the three quality registries, were recruited from one hospital participating in the survey. The clinic was selected based on a convenience sampling. The physicians, nurses and managers involved with each NQR at the particular clinic were contacted via email. Email addresses were identified either
25 Ann Catrine Eldh et al., “Factors Facilitating a National Quality Registry to Aid Clinical Quality Improvement,”
Submitted, n.d., 27.
via phone calls to the clinic, or through a list of emails attained within the larger research project.
At this point, oral information about the study was provided and the engagement in the NQR was confirmed. Further, all potential informants received an email with information about the study and they were asked to participate in an interview. Approval by writing a reply agreeing to participate in an interview was considered consent. A total of six persons consented and participated in individual interviews.
Ethical approval for interviews and the survey was obtained from a Regional Ethical Review Board prior to the study (Uppsala 2013/181).
Data collection and measures Questionnaire
The questionnaire was based on prior studies and literature reviews 26 that had exposed several factors for further investigation. The questionnaire was distributed in Swedish and the complete survey can be obtained from the research team. 27
The research team developed the survey exclusively for this study, including content validity and response process validity with regards to structure, content and layout28. The final version of the survey comprised 50 questions. These were organized as: (A) Demographics; (B) Perceived quality of the care provided; (C) Quality of registry data; (D) Organizational conditions; (E) The respondent’s use of registry data; (F) The unit’s use of the registry data, and (G) Perceived value of the registry.29 The same survey was used for all three registries, with only differences in the name of the NQR. For most items, a Likert scale was used for the responses, with an opportunity to provide additional remarks in free text to each section.30 The quantitative data is reported in a series of papers31, while this report focuses on the free text responses; these responses had not been analysed before. All in all, there were 127 individuals who submitted 275 free text responses to the survey. The distribution from respective registry can be found in the two subsequent tables.
Table 1 demonstrates the distribution of free text responses where Riksstroke was quantitatively
26 Eldh et al., “Facilitators and Barriers to Applying a National Quality Registry for Quality Improvement in Stroke Care”; Mio Fredriksson et al., “Local Politico-Administrative Perspectives on Quality Improvement Based on National Registry Data in Sweden: A Qualitative Study Using the Consolidated Framework for Implementation Research,” Implementation Science 9, no. 1 (December 2014), doi:10.1186/s13012-014-0189-6.
27 Eldh et al., “Factors Facilitating a National Quality Registry to Aid Clinical Quality Improvement.” Submitted, n.d., 27
28 Ibid.
29 Ibid.
30 Ibid.
31 Ibid.
largest with 146 responses of the original total of 275 responses. From all three registries 199 free text responses were used for analysis. Disqualified free texts responses were those not providing anything to the analysis, for example those directing the investigator to another question for information. The total amount of responses from the six interviews that corresponded to
subcategories or categories from the analyze matrix are also demonstrated in Table 1, separated by respective registry.
Each participant in this study submitted a minimum of one response to the questionnaire in total and a maximum of one response per questionnaire section. The section covering quality of care generated the most free text responses and the complete distribution of responses between sections is demonstrated by Table 2. The section about the demographics of the respondent is not included in the analysis and is therefore not demonstrated by Table 2.
Table 1. Distribution of free text responses and interview responses.
NQR
(Nation Quality Registry)
Riksstroke (amount)
GallRiks (amount)
Nationellt
Lungcancerregister (amount)
Total (amount)
Free text responses originally
146 88 41 275
Free text responses used for analysis
115
~58 %
48
~24 %
36
~18 %
199
Individual respondents 64 46 17 127
Amount of identified subcategories from analyze matrix to interviews
29 14 17 60
Amount of identified individual categories from analyze matrix to interviews
13 7 8 28
Table 2. Distribution of responses from survey section and NQR Original free text
responses from:
Riksstroke
A total of 64 individuals left one or more comments distributed over these sections:
GallRiks
A total of 46 individuals left one or more comments distributed over these sections:
Nationellt
Lungcancerregister
A total of 17 individuals left one or more comments distributed over these sections:
All three NQRs total
Survey section:
Quality of care
31 17 9 57
Data quality 23 20 9 42
Organizational conditions
30 11 5 46
The
respondent’s use of registry data
27 12 7 46
The stroke/
gall stone/lung cancer unit’s use of registry data
16 11 4 31
Perceived value of the registry
19 17 7 43
Interviews
Three interviews were conducted face-to face in a semi-structured manner in a hospital setting, location selected by the participant. These three were accomplished by the author of the article.
The other three interviews were carried out by another investigator through phone calls, date and time selected by the participants’ preferences. Each interview lasted under 45 minutes and used the same semi-structured interview guide, as presented in Attachment A (developed by the research group).32 All participants had the possibility upon request to obtain a copy of the interview and revise information given.
The questions in the interview guide regarded working with a quality registry and assessed: the participants role in the registry, benefits and disadvantages with the registry, the kind of quality improvement work that was conducted from registry data, the participants thoughts about the
32 Eldh et al., “Facilitators and Barriers to Applying a National Quality Registry for Quality Improvement in Stroke Care.”
different variables in the registry, resources for registry participation, time allotted to registry work, usage of registry data, request of registry data from a third part, correlation between the registry and national guidelines for the particular health problem, sources of the evidence used for treating the particular health problem and thoughts from the participants before or during the interview. Example questions regarding benefits and disadvantages with the registry included
“What kind of benefits and disadvantages do you see with the registry and/or your participation in the registry?” Questions regarding resources for registry participation included “What kind of resources do you have for registration/participation in the registry? Resources concerns both financial resources as well as resources such as staff and other support.” If not spontaneously brought up the investigator would also ask “Do you get any support for participating? If so, where from?” As such, the interviews provided for a more in-depth understanding of the experience of health care professionals working with quality registries in a clinical setting. Each interview was audio recorded and later transcribed verbatim to text files.
Data analysis
The voluntary comments from the questionnaire were analyzed by content analysis33. All data was first read several times to gain a naïve understanding of the text, as a whole. Subsequently, free text comments with related content were assembled, to provide for coding. Each sentence was condensed, and coded. The codes, and their corresponding sentences, were formed as
subcategories, successively formed into 15 categories. The data analysis was made by the author, under supervision of a senior member of the research team, with several years of experience using content analysis and of health care quality research.
After the composition of subcategories and categories from the free text comments in the survey, these were used as a matrix for analysis of the interviews using a deductive approach.34 Thus, the texts from the interviews were read from beginning to end, and units corresponding to the
subcategories and/or categories were incorporated in the analysis. That is, each part of or entire sentence, paragraph or passage was marked to identify which subcategory and/or category it corresponded to, and included in the analysis to further inform the subcategories and categories.
To conclude, the categories were associated in overarching themes.
33 Satu Elo and Helvi Kyngäs, “The Qualitative Content Analysis Process,” Journal of Advanced Nursing 62, no. 1 (April 2008): 107–15, doi:10.1111/j.1365-2648.2007.04569.x.
34 Ibid.
Results
In the present study free text responses from 127 individuals to a questionnaire over three NQRs were assed along with 6 individual interviews as presented in the method. The categories are presented below, with italicized text to illustrate the subcategories of each category. A total of 15 categories were formed by the analysis and can be found, along with the subcategories, in
Attachment B. All 15 categories are also presented with the overarching themes in Figure 1. The first theme represents present utilization and opinions about NQRs, the second represents
identified barriers to optimal utilization and the third represents available support and suggestions for improvement, i.e. facilitators.
Figure 1. The 15 categories acquired from content analysis classified barriers, facilitators and after present utilization.
Quality improvement and
NQRs today
The organization’s application of
registry data
The registry provides a foundation to
quality improvement
Good results in the registry can always
be better
The organization’s unsatisfactory application of registry data
The registry does not provide the main foundation to
quality
Barriers
Insufficient resources affect
the quality of health care
The reliability of registry data varies
The registry’s coverage varies
Delayed feedback on registry results
Deficits in registry structure
Facilitators
Help and support primarily locally or from the NQR
Registration and management of
registry data require specific
competence
Registration and improvement work
require allocated resources
Clarified incentives encourage participation
Commitment from several professions
facilitate application of
NQRs
THE ORGANIZATION’S APPLICATION OF REGISTRY DATA
Requests of registry data came mainly from the own organization or the NQR. Registry data had also been presented via media to reach out with public information. Registry data was used to identify quality indicators, to review quality indicators, for feedback, in research, in budget negotiations and overall to access different types of information for an organizational overview.
There was also a report of an increased interest of registry data from the pharmaceutical industry.
Registry data also increased an interest of comparison between health care units which many reported as a positive influence; an instrument promoting equal health care regardless of location.
Comparison also facilitated questioning and reflection of organizational methods and techniques.
How data was accessed differed from independent withdrawal of information to relying on other people to obtain, analyze and report data. Nurses were frequently reported to be involved in registry administration along with physicians whereas utilization and benefits of an NQR was described belonging foremost to physicians. Despite no reported individual usage of an NQR most nurses still described an NQR as something positive.
“Then there’s the evaluation of data… [---] there’s an incentive when you see that… there are this many clinics doing it this way, why aren’t we doing that?” (6 [interviews], GallRiks)
THE REGISTRY PROVIDES A FOUNDATION TO QUALITY IMPROVEMENT
Registries have initiated most of the quality improvements and have highlighted needs of improvement according to many. Many stressed the need of an NQR as an absolute necessity to have any indication about quality and results from the organization while others thought that registration gave little to a small scale organization although it was of importance in a holistic perspective. There was always something to improve and therefore NQRs were of assistance.
NQRs could help visualize deficient parts of an organization as possibilities to improve. It also gave a lot of information where Sweden was seen rather unique internationally.
“Well, I do put a lot more time than others on registration. But it is still so that I […] I actually think it is, it is worth it. […] it adds something that you are glancing on the… quality. And it is an
incentive too…” (6 [interviews], GallRiks)
GOOD RESULTS IN THE REGISTRY CAN ALWAYS BE BETTER
Most who reported good results also reported room for improvement. The registry helped to visualize improvement, both achieved and what still could be worked on.
“Some results are good, but we have a lot left to work with and improve.” (209 [questionnaire], Riksstroke)
THE ORGANIZATION’S UNSATISFACTORY APPLICATION OF REGISTRY DATA
Whereas many areas of application were described by study participants these were sometimes insufficiently utilized. Employees responsible for quality didn’t have the ambition to utilize NQRs and many wished to forward more registry data. Some stated that no information was collected from the NQR and others that they didn’t know how physicians utilized registry data. Registry data was said to be more informative than what the clinic currently took use of but it had to be up- to-date to be utilized optimally.
“[…] and these quality registries are very important because you’re collecting a lot of information. […] but I can’t say that we have found the dead on target model, I don’t think so. I
still think that we are kind of fumbling forward. […]” (2 [interviews], Riksstroke)
THE REGISTRY DOES NOT PROVIDE THE MAIN FOUNDATION TO QUALITY IMPROVEMENT
Some claimed that the majority of quality improvements had come about irrespectively of an NQR. There were fewer comments under this category than its counterpart but is likewise represented.
“We don’t […] work with quality improvement from RS [Riks-Stroke].” (138 [questionnaire], Riksstroke)
INSUFFICIENT RESOURCES AFFECT THE QUALITY OF HEALTH CARE
Comments from all three NQRs stated that there’s a lack of resources negatively affecting health care. Improvement work needed to be prioritized and directors didn’t allot time for improvement work. Lack of staff limited the organization, reduced time for patient appointments, increased tempo and resulted in a high employee turnover that aggravated quality improvement. Shortage of physicians meant an irrational responsibility for inexperienced physicians. Lack of resources in the organization meant an insufficient capacity for specialized health care and treatment of stroke patients outside of specialized stroke units as well as insufficient time lots for surgery that didn’t cover the actual need (for gall bladder surgery). Insufficient resources in several areas meant longer (malignancy) investigation processes.
“However, the resources required to fully accomplish the improvement work, which registry result shows necessary, does not exist.” (26 [questionnaire], Riksstroke)
THE RELIABILITY OF REGISTRY DATA VARIES
Equal registration was important to achieve trustworthy data and one major concern from study participants. Few considered registry data as comparable between units because of difference in specificity and likeliness of understatements of complications. There was no national consensus or unanimity of how data should be reported and therefore it wasn’t possible to assume that all units registered correctly. Authorities were also thought to impact documentation and interpretation of results and trust to registry data increased higher up in the organization. Apart from differences in registration there were also problems with differences in health care as for example: amount of patients, regional uptake numbers, specialized care and length of follow ups.
“The problem with registries is that all units don’t document in an equal way and don’t have the same assessment when it comes to postoperative complications and then it gets hard to compare
complications between different units.” (11 [questionnaire], GallRiks)
THE REGISTRY’S COVERAGE VARIES
The variation of coverage was only addressed by professionals working with the stroke NQR.
People reported either a good level of coverage or an insufficient one. Coverage could suffer from incorrect diagnosis and missed diagnoses.
“How many of the stroke diagnoses are really correct? I have unfortunately seen numerous patients diagnosed with stroke from colleagues when the real diagnosis was something entirely different! How many stroke diagnoses have then been missed?”(79 [questionnaire], Riksstroke)
DELAYED FEEDBACK ON REGISTRY RESULTS
The time between registration and access of data was considered too long. These delays prevented usage of registry data in everyday work and there were also complaints of deficient regular
reports and analysis from the NQR centre. There had however been an improvement recently.
This category was unique for the lung cancer and stroke NQR.
“The data that is accessible has such a delay that it can’t be trusted.” (9 [questionnaire], Nationella Lungcancerregistret)
DEFICITS IN REGISTRY STRUCTURE
Many deficits were identified as well as many suggestions for improvement regarding structure.
One mutual problem was the impossibility to motivate reported actions, where some actions were misplaced and misjudged due to this. Other problems were: no possibility to see specific interests in the registry, absence of questions regarding nursery and patient experiences, no consideration to differences in size between reporting units, unclear questions making significant differences in interpretation and no consideration to complicating factors. Many wished for integration with medical record systems and registries, reducing errors and saving time and staff. Another concern was outdated variables and non-validated variables reducing credibility. Interviewed participants described minor problems and suggestions for improvement but were, however, in general content with the registry structure.
“Some collected information is vague and is therefore interpreted incorrectly. […] There are only yes or no alternatives. No is interpreted adversely but there might be medical reasons for it. But
that [information] is not requested. There are also other examples.” (49 [questionnaire], Riksstroke)
HELP AND SUPPORT PRIMARILY LOCALLY OR FROM THE NQR
For registry related questions most participants directly addressed the registry centre itself and were usually satisfied with the help offered describing it as swift and convenient. Locally it varied more and some turned directly to the NQR centre out of convenience, not knowing if there existed any local support. This however applied mostly to those in charge of registry administration, others turned to the resource person assigned to registry administration or expressed a need of such a resource. Few participants knew if there was any financial compensation for participating in the NQR.
“They [NQR employees] are very easy to get a hold of and they are responding quickly to emails and you are directly referred to the right person.” (3 [interviews], Riksstroke)
REGISTRATION AND MANAGEMENT OF REGISTRY DATA REQUIRE SPECIFIC COMPETENCE
To evaluate quality there was a need to be involved in the actual health care. Furthermore it was meaningful with a specific employment covering handling of everything registry-related to have time, competence and experience needed. Management of registry data was reported to be difficult and those who only registered a small number of times usually needed repeated instructions.
Many also reported that the NQR only had an application to the work of physicians.
“It is crucial to have people responsible for registration with time and knowledge about registries”. (32 [questionnaire], Riksstroke)
REGISTRATION AND IMPROVEMENT WORK REQUIRE ALLOCATED RESOURCES
Time was considered important for registration and analysis to yield something. There was insufficient time for own usage, insufficient time for registration and no time allotted for analysis or registration. It was likewise important that exchange from the registry corresponded to the time spent on registration. Registration required, apart from time, an allocated employment that wasn’t person-bound. Lack of staff made it hard to accomplish registration in time. Physicians had a shortcoming in registration and needed administrational support.
“I just wished I had more time” (165 [questionnaire], Riksstroke)
CLARIFIED INCENTIVES ENCOURAGE PARTICIPATION
There was a request of a more articulated benefit of an NQR to encourage participation. Rewards in some form should be given to those with exemplary participation, for example with good registration of follow-ups on patients.
“Yes, some might see it as a burden […] one should sit down and fill in a lot of difficult forms, which you don’t find any value in.” (1 [interviews], Riksstroke)
COMMITMENT FROM SEVERAL PROFESSIONS FACILITATE APPLICATION OF NQRS
The interest for NQRs was generally low and many wished that it would increase. It was
recognized as a good tool for quality improvement but it also required involvement of managers and leaders. If many people weren’t involved and interested it would easily fail. To increase understanding and interest simple relationships between NQRs and quality improvement had to be presented and adjusted to include different employment types.
“Yes, absolutely because we are helping each other. Yes, it is of importance. […] Because it would easily fail if you were alone.” (3 [interviews], Riksstroke)
Summary of results
In summary, the results showed that limited resources in health care affected not only the quality of health care but also quality improvement work; even if registry results revealed possible
improvement areas there were insufficient resources to fully complete them. The current use of registry data includes reviewing of quality indicators and other areas of application that could be related to quality improvement: for comparison between units, for feedback and for research.
Despite various areas of application depicted and many professionals involved in registries, physicians were the only ones who stated having any use of the registry – even though nurses and others reported NQRs as something generally positive. Not only resources showed to limit the application of registries, some employees lacked ambition to utilize it, a few stated that no
information was collected from the NQR and others thought that NQRs wasn’t yet utilized to full capacity. One suggested way to increase the application of the NQRs was to increase the
commitment of others: managers and leaders were pointed out as particularly important. This required for those already engaged to have easily comprehended results at hand. The benefit of an NQR also had to be emphasized to increase application and some suggested rewards for those units whose participation were exemplary in means of high percentage of registration and complete documentation for instance. For complete documentation allocated registry staff was identified as essential; they need time, competence and experience to facilitate management of registry data. Staff who only registered once in a while were considered prone to forget procedures and to need frequent guidance. A protruding example of insufficient resources that diminished registration and improvement work was time: time for registration and analysis, and for one’s own use of data,. Another barrier correlated to time was the long delay in registration; this was
particular for fellow staff who did not have a specific assignment for a registry or where some registration had to be completed by various individuals, like post-surgery. Help and support for registry work was primarily sought and attained from registration administrators in the particular clinic. Those assigned to registry work primarily turned to the NQR centre for support and were usually pleased with the support and assistance offered.
Discussion
Regardless of the obstacles many stakeholders reported good results in their registry but also stressed that there was a lot left to improve in terms of quality aspects of care, which the registries had highlighted. Quality improvement was implemented by using the the registries although a few thought improvements came irrespectively. Some of those in favor of NQRs even thought that a registry is an absolutely necessary tool in quality improvement work even though insufficient resources constituted a barrier.
Value of registries for health care
Eldh et al35 found that NQRs could have a value for quality improvement in health care given certain stipulations. One of these was a leadership engagement, which corresponds to the
statements of participants in this study requiring managers and professionals in leading positions to carry out improvement work. The other identified key factors focused more on means of quality improvement alone, i.e. developing the stroke process and structures for mutual learning, rather than general utilization of an NQR for quality improvement, and these are therefore not well-suited for comparison. Fredriksson et al36 found that despite an overall positive view of NQRs their data was not used for development and improvement. This was mainly due to immature meso-level infrastructures and processes on a politico-administrative level. Instead, other data was prioritized and used, like regional comparisons. This differs to the overall result in this study of registry data implementing most of the improvement work in health care, not
specified on which organizational level, but still recognized by participants as a fundamental tool for quality improvement. A supposed value of NQRs are however recognized in the study by Fredriksson et al but obstacles like reliability in data, latency of data and insufficient resources to analyze and utilize data are addressed all in line with my present findings. Outside of Sweden, Black and Tan (in the UK) found that national clinical registries could have a value for usage, although the application of databases was scarce37. Where used, registries indicated a value and were therefore suggested to be an under-utilized source. This is in line with the findings in this report, and Black and Tan argue for an apparent usefulness: i.e. comparison between units, evaluating provided health care and access to descriptive information, but there’s a lack of utilization. Once again, the reasons behind that reflect some of the ones found here in my study:
like how the professionals in the UK not being aware of the potential use corresponds to how participants in this study require more pronounced benefits to increase application. Access to data is also addressed as a problem where the data often is readily available but there’s a misconception of a requirement of requested access and permission. The Swedish NQR participants of this study describe accessing data as difficult due to structural reasons; they know where but not entirely how to access data. Coverage and quality of data remains a shared concern and emphasize the need of reliable data for professionals to value NQRs as a tool for improvement of health care.
35 Ann Catrine Eldh et al., “Depicting the Interplay between Organisational Tiers in the Use of a National Quality Registry to Develop Quality of Care in Sweden,” BMC Health Services Research 15, no. 1 (June 2015),
doi:10.1186/s12913-015-1188-2.
36 Fredriksson et al., “Local Politico-Administrative Perspectives on Quality Improvement Based on National Registry Data in Sweden.”
37 Black and Tan, “Use of National Clinical Databases for Informing and for Evaluating Health Care Policies.”
Differences in perception of quality registries between professionals
There seems to be a difference in how professionals regarded NQRs: Nurses were involved in documentation but had no use of it in their clinical work. Still, they regarded NQRs as something positive. They thought that the registry would benefit the patient in the end and therefore
considered it valuable. Physicians were considered the only category to have an application of registry data and those who should carry it out. Physicians thought that it served quality to review actions and results but they had many concerns about the validity of data. This was an immense obstacle. Managers believed more in data and registry results despite modest understanding of practical documentation and registry work. They regarded NQRs as a good tool for quality improvement and they thought that reviewing data was significant. Administrators or secretaries described a tough work load due to registries and no desire to increase it further and when asked about the benefit they would direct the question to a physician. Conclusively, nurses saw a benefit regardless of no own usage, administrators regarded it more as work, physicians were cautiously positive and requested certain fulfillments before admitting certain value and managers were positive despite limited knowledge of application. Managers’ trust to data and a limited insight to the NQR is also in line with the findings of Eldh et al38 who found that professionals on all levels of health care see registries as an administrative constraint and cost. Naturally, some differences can be detected in prior studies39 but these primarily owe to the fact that they consider areas of application for one specific profession. Managers cannot be expected to answer questions
regarding documentation since they are not involved, nurses can most likely not answer questions of financial compensation for participating in registries etcetera. Opinions about registry data, coverage and the organization around registries are addressed in prior studies but the overall perception of an NQR out of professional category is found to be a new area for investigation.
While the demographic features of the participants of this study are not presented to assure anonymity the addressed differences in perception of NQRs can be distinguished from the
subcategories. The distinguishable difference is interesting since no prior study has reflected on it.
Is this another aspect of implementing improvement work from quality registries? Would an involvement of nurses, who seem to have a more optimistic view of NQRs, be desirable to increase implementation? Are their roles in the registries connected to their perception of it and could they be modified to better carry out improvement work?
38 Eldh et al., “Depicting the Interplay between Organisational Tiers in the Use of a National Quality Registry to Develop Quality of Care in Sweden.”
39 Ibid.; Fredriksson et al., “Local Politico-Administrative Perspectives on Quality Improvement Based on National Registry Data in Sweden.”
Value of resources for NQRs
Another core finding is the recurring issue of resources. Insufficient resources appeared in both data sources and all contexts and were considered a protruding barrier for improving health care, using quality registry outcomes. While Black and Tan40 reviewed the applications of 15
established national clinical databases in England, insufficient resources were not considered one of the main barriers to utilize them. This is contradictory to my findings where there was a strong notion of insufficient resources to carry out the improvement possibilities which NQRs reveal.
Prior studies in Sweden have, on the other hand, found resources to be an important part in quality improvement with regards to NQRs. The fact that documenting and validating data consumes a lot of time and resources is supported by my findings, but also holds true for the implementation of results; there is not sufficient time left for outcome analysis which removes the bridge to
improvement41. Lack of time, and more specifically staff, are acknowledged both at the clinical level but also at a meso-level, i.e. by politicians and administrators42.
Aims and results
This study intended to broaden the knowledge of the relationship between quality improvement and NQRs and such connections have been presented in the result and discussion. One of the aims with the report was to investigate which factors professionals described to affect quality
improvement related to quality registers in terms of certain topics and these factors can be found in results and in Attachment B. The other question was whether or not there was any difference in the factors physicians describe as obstacles and facilitators for quality improvement to their patient groups, in relation to quality registers, compared to those described by managers and other
professionals. This has been discussed, primarily from the perspective of perception of NQRs between professionals. Otherwise; such differences can be distinguished primarily in the subcategories. Regarding obstacles and facilitators there were no protruding differences. For example, the recurrent complaint of data, trustworthiness and coverage, were mutual concerns by all employment categories.
Methodological considerations
The data from this study is of a qualitative nature and have therefore been analyzed by qualitative methods. One of the apparent risks is that of subjective results despite following the rules behind deductive content analysis. The flexibility of content analysis is one of its advantages but also
40 Black and Tan, “Use of National Clinical Databases for Informing and for Evaluating Health Care Policies.”
41 Eldh et al., “Depicting the Interplay between Organisational Tiers in the Use of a National Quality Registry to Develop Quality of Care in Sweden.”
42 Fredriksson et al., “Local Politico-Administrative Perspectives on Quality Improvement Based on National Registry Data in Sweden.”
poses a potential threat to validity; less standardization and fixed directions make it challenging for researches to draw the same conclusions due to differences in interpretation43. The results are in other words highly dependent of the investigator and its analytic abilities. 44 The method is, however, suitable for quantitatively large amounts of qualitative data as interviews or other multifaceted textual data. 45 However, this study benefits from being part of a research project were similar data is investigated and analyzed both by qualitative methods and quantitative
method and thus results from different investigators could be compared and questioned. This study has been limited to the free text comments from a national survey of three NQRs and one
consideration is whether these three could represent NQRs in general, currently 96 registered46. More are, however, included in the national research project carried out by three different research groups. National quality registries naturally share some qualities to be considered an NQR but there are still differences between them that might affect the results. Riksstroke, for example, is a well-established registry initiated 199447 while GallRiks is much younger and started in 200548. This might also account for the apparent differences in responses between the registries in this study. Riksstroke is the quantitatively most represented, and these differences should be taken into consideration. Two of the registries included are also foremost internal medicine registries,
Riksstroke and Nationella lungcancerregistret, were GallRiks is primarily surgical. However, both are medical registries of importance to patients, professionals, health care organizations and society. Given recognition of particular issues with single registries, facilitators and barriers to quality improvement could supposedly apply in different contexts. Applying and comparing strategies from different areas is, after all, nothing new. Health care has, for example, previously applied safety strategies from the aeronautical industry49.
Six interviews have been included to enhance analysis. This is realistic given the type of study, even though a larger study could have benefited from more interviews, particularly from a wider
43 U.H Graneheim and B Lundman, “Qualitative Content Analysis in Nursing Research: Concepts, Procedures and Measures to Achieve Trustworthiness,” Nurse Education Today 24, no. 2 (February 2004): 105–12,
doi:10.1016/j.nedt.2003.10.001.
44 Elo and Kyngäs, “The Qualitative Content Analysis Process.”
45 Ibid.
46 Christensson, “Nationella Kvalitetsregister - SKL.”
47 Riksstroke, “Allmän Information,” Allmän Information |, accessed March 1, 2016, http://www.riksstroke.org/sve/omriksstroke/allman-information/.
48 GallRiks in collaboration with UCR, “Om GallRiks,” Om GallRiks- GallRiks, October 7, 2015, http://www.ucr.uu.se/gallriks/index.php/om-gallriks.
49 Elisabeth Haddleton, Teamarbete Och Kommunikation - Crew Resource Management - CRM - Vårdhandboken, Vårdhandboken (Sveriges landsting och regioner, 2015), http://www.vardhandboken.se/Texter/Teamarbete-och- kommunikation/Crew-Resource-Management-CRM/; Läkartidningen, “Sjukvården Och Flyget Har Mycket Att Lära Av Varandra I Säkerhetsarbetet,” Läkartidningen Nr 10 2008 Volym 105, no. 10 (2008): 743–44.
variety of professions. To protect the integrity of the interview participant their answers have been included with the free text comments in the result table, and their professions and/or clinics have not been revealed. The object of analysis is still extensive, aiming to broaden the knowledge of the relationship between quality improvement and NQRs, especially in relation to the parallel studies by the research group.
Conclusion
The results in this study show that validity of data is a strong factor for participants’ confidence in and utilization of data. Other factors are also of importance, such as: regular presentation of results and an integration of registry work in the everyday clinical practice. However, presenting data that the staff, that is, professionals who the registry is intended to benefit, does not trust is ineffective, and jeopardize the use of data for quality improvement work. Validation and national consensus of registry structure, questions and documentation should therefore be a top priority.
Lack of time and resources is also a protruding finding in this work: registry work can become more effective both structurally and technically and an allocated assignment for registration of data seems to be effective. The findings in this paper suggest that registries support quality improvement work but that their full potential is far from utilized. The amount of time spent on registry work need to justify the benefit of them, as emphasized by this study’s participants.
Further, there are many suggestions and ideas of how to improve NQRs so that they in return could help improve quality in health care. The findings mainly confirm the results of previous studies50 but they also add the perspective of potential benefits of comparing different professions and roles vis-a-vis registries. The findings of this alongside other studies have a potential to support a developed and improved utilization of a quality registry in today’s health care, to subsequently improve quality of care. Most importantly, these findings argue for an increased focus on quality improvement using NQRs, and time allocated for professionals involved in clinical registry work to facilitate this process.
Acknowledgments
I would like to thank my supervisor, Ann Catrine Eldh, for patient guidance, encouragement, interview data and advice with all aspects of this study and research article. Thanks go to the Department of Public Health and Caring Sciences for the use of questionnaire data and valuable feedback from its research group.
50 Bradley, “Data Feedback Efforts in Quality Improvement”; van der Veer et al., “Improving Quality of Care. A Systematic Review on How Medical Registries Provide Information Feedback to Health Care Providers.”
