by
Pieter Nomden
University of Groningen Faculty of Economics and Business
MSc Business Administration
Specialization Operations & Supply Chains August, 2009
Aquamarijnstraat 575 9743 PN Groningen
06-54253257 p.nomden@student.rug.nl
s1654918
Organization: Universitair Medisch Centrum Groningen Company supervisor: Anja Boot
RUG supervisor: Dr. Manda Broekhuis RUG second supervisor: Dr. Ir. D.J. van der Zee
PREFACE
Dear reader,
ten years, two lustrums, or one decade. After two years at the MBO and five years at the HBO, I started with the so-called bridging program of the Faculty of Economics and Business in September 2006. And now - almost three years later - I am writing the preface of my mas- ter thesis. The completion of this thesis will be the end of my life as a student too. Though I can not end the chapter Orthopedics at the UMCG by now, not yet. Since I received a serious injury on my knee during a football match, they have not seen the last of me.
Many people supported me during the process of writing this thesis. First of all, I would like to express my thanks to my supervisor at the UMCG. Anja Boot was always available to an- swer my questions, to explain the medical terminology, and to point me in the right direction.
Furthermore, I would like to thank my roommates Corian, Jantina, and Imke. Corian provided me with the necessary distraction with her stories on wind mills, her pets, and the problems with her daily activities. Jantina was a fellow student of Business Administration and it felt good to have someone around who did not understand the medical terminology as well. And with Imke I had numerous, amusing discussions about the introduction of economic frame- works in the medical environment. If I remember correctly, I could never let her agree with me.
My supervisor from the University of Groningen was Manda Broekhuis. During the writing of my master thesis, she provided me with a lot of critical and useful advice. Furthermore, she supported me when I struggled to finish this thesis. In the end, she spend a lot of time on me – even more then could be expected of her – and I can not thank her enough for that. In the final stage of writing this thesis, Durk Jouke van der Zee provided me with additional comments which enabled me to make the finishing touches. I would like to express my thanks for his work.
Finally, I would like to thank my family and friends for their support in the preceding year.
They provided me with advice, distraction, and – eventually – social pressure to finish this thesis. And though I might not say this every day, but I love you guys.
Pieter Nomden,
Groningen, August 14th 2009.
SUMMARY
The aim of this research is to identify starting points for further improvement of the treatment of elderly patients with a hip fracture. After the identification of these starting points, the ac- tivities which lead to a better performance are elaborated.
The basis of this research is the following research question:
Which changes in the present organization and planning of elderly patients with a hip frac- ture are necessary, in order to provide these patients with the right care i.e. the care with a minimum lead time and the best possible quality of life?
With the use of a theoretical framework, the relationship between the different variables and the outcome measures are analyzed. These are visualized in the figure below.
As the figure shows, the influence of the patient, demand, and treatment characteristics has a large influence on the planning issues and subsequently the outcome measures. The perform- ance of the UMCG on survival and total lead time is above average compared with other UMC’s. Furthermore, none of the actionable variables for the outcome measures is unambi- guous. This shows that presence of mortality, complications, secondary surgery, and the length of the lead time can not be influenced by the actions of the UMCG.
4. Planning characteristics
1. Patient characteristics
3. Treatment characteristics 2. Demand
characteristics
5. Capacity 6. Criterions
7. Outcome measures
The results lead to a number of recommendations derived from this research:
o A stronger focus on the discharge procedures will reduce the non-essential lead time after surgery and improve the cost-effectiveness of the department.
o The UMCG should make use of planned emergency surgery in the emergency operating rooms of the UMCG and cluster the patients at one department in order to improve the quality for the patient.
o The UMCG should perform surgery for this diagnosis within one calendar day after ad- mission in order to reduce the presence of complications due to the fracture.
o It is necessary to perform a prospective follow-up study to gain an in-depth knowledge on the outcome measures – both of this study and for future patients with a hip fracture.
o In order to perform this follow-up study, it is necessary to register the data
PREFACE ... 2
SUMMARY ... 3
1 INTRODUCTION... 7
1.1 Problem Statement ... 7
1.2 Goal ... 8
1.3 Research Question... 8
1.4 Structure ... 9
2 SYSTEM DESCRIPTION ... 10
2.1 Organization ... 10
2.2 Routing of the patients ... 10
2.3 Planning issues ... 12
2.3.1 Admission issues... 12
2.3.2 Surgery issues... 13
2.3.3 Discharge issues... 14
3 THEORETICAL FRAMEWORK ... 15
3.1 Theoretical framework ... 15
3.2 Patient, demand, and treatment characteristics ... 16
3.3 Survival ... 18
3.4 Quality of life ... 20
3.4.1 Complications due to the fracture... 20
3.4.2 Secondary surgery... 21
3.5 Lead time... 22
3.5.1 The need for lead time reduction... 23
3.5.2 Lead time before surgery... 24
3.5.3 Lead time at surgery centre... 24
3.5.4 Lead time after surgery and the total lead time... 25
3.6 Admission planning... 26
4 METHODOLOGY... 28
4.1 Research strategy... 28
4.2 Research population ... 28
4.3 Data collection and variables ... 28
4.4 Data analysis ... 31
5 RESULTS... 33
5.1 Descriptives of the research population ... 33
5.1.1 Characteristics of the research population... 33
5.1.2 The composition of the lead time in the UMCG... 35
5.2 Mortality... 36
5.3 Complications due to the fracture ... 37
5.4 Secondary surgery ... 38
5.5 Lead time issues ... 39
5.5.1 Lead time before surgery... 39
5.5.2 Lead time at surgery centre... 41
5.5.3 Lead time after surgery... 42
5.5.4 Total lead time... 43
5.6 Planning issues ... 45
5.6.1 Admission... 45
5.6.2 Surgery... 47
5.6.3 Discharge... 49
6 DISCUSSION ... 51
6.1 Mortality... 51
6.2 Complications due to the fracture ... 52
6.3 Secondary surgery ... 53
6.4 Lead time issues ... 54
6.4.1 Lead time before surgery... 55
6.4.2 Lead time at the surgery centre... 56
6.4.3 Lead time after surgery and the total lead time... 57
6.5 Planning issues ... 59
6.5.1 Admission planning... 59
6.5.2 Surgery planning... 61
6.5.3 Discharge planning... 63
7 CONCLUSION ... 65
7.1 Conclusions ... 65
7.1.1 Impact of patient, demand and treatment characteristics on the outcome measures and the planning of surgery and discharge... 66
7.1.2 Impact of planning and resources on outcome measures... 67
7.1.3 The actionable variables which can improve the performance of the UMCG. 68 7.1.4 The answer to the main research question... 69
7.2 Recommendations ... 69
7.3 Further research... 70
APPENDIX 1: The influence of variables on the mortality within one year after surgery …..72
APPENDIX 2: The influence of the variables on the complications due to the fracture ……73
APPENDIX 3: The influence of the variables on secondary surgery within one year ………74
APPENDIX 4: The influence of the variables on the lead time for intercapsular fractures …75 APPENDIX 5: The influence of the variables on the lead time for extracapsular fractures ...78
1 INTRODUCTION
Like most western countries, the population of the Netherlands is subject to obsolescence. In 2007, 14% of the population was a senior citizen – 65 years of older – and this number will increase to 24% in 20501. These developments in the demographical composition in the popu- lation lead to changing needs for the health care industry. Therefore, increased attention exists for the care and support for elderly patients.
In 2004, the three-year Healthy Ageing project started with their investigation of the available data on both health and ageing. The definition and purpose of this project were described as
“… the process of optimizing equal opportunities for health to enable older people to take an active part in society and to enjoy an independent and good quality of life.”2. The final report contains the recommendations for the care of elderly patients – aged 50 or older – and the strategy for implementation.
In line with the goals of the Health Ageing project, the University Medical Center Groningen – UMCG – started multiple projects to improve the care for elderly patients. One of them in- volves the development of a clinical pathway for patients with a hip fracture of 60 years and older which measurable improves the survival and the quality of life of these patients. This study will provide an analysis on this subject of this population in the years 2006 and 2007. In addition to this, it will investigate the lead time in the hospital and the planning issues at the hospital.
1.1 Problem Statement
The prevalence of age related diseases like hip fractures will increase – 20 % of the women and 10% % of the men at age 50 will at some point of their life suffer from a hip fracture – and it is essential to anticipate contraceptive. As a result, it will be necessary to use the avail- able resources more efficiently, to create more capacity or both.
In the present situation, the number of admissions is dependent on the available capacity of hospital beds. The utilization of this bottleneck capacity is therefore of the highest priority.
Since the shortening of lead time implies a more efficient use of beds, the lead time per pa- tient is determined and analyzed for improvements. Subsequently,
This project leads to a number of questions with relation to the actionable variables – those factors which are amenable to change – which influence the outcome of the treatment. The outcome of the treatment involve qualitative measures – survival and quality of life – and
quantitative ones – the lead time in the hospital. In an attempt to support the department with the creation of a clinical pathway, this study focuses on the identification of the significant variables in relation to these outcome measures.
1.2 Goal
The objective of this study is to provide insights for the reduction of the lead time for patients with a hip fracture of 60 years and older. This should be established without any negative consequences for the quality of life of these patients. Therefore, the outcomes of the treat- ment with relation to the qualitative aspects – survival and quality of life – are subsequently investigated. The present performance on both quantitative and qualitative aspects is unknown and requires further investigation to allow purposeful improvements.
1.3 Research Question
This results in the following main research question:
Which changes in the present organization and planning of elderly patients with a hip frac- ture are necessary, in order to provide these patients with the right care i.e. the care with a minimum lead time and the best possible quality of life?
The answer to this research question will provide the organization with the knowledge of their present performance, the possibilities for improvement, the relative impact and expected re- sults on these changes on the processes, and the starting points for further research.
A set of sub questions was defined to answer the main research question, as mentioned above.
The answers of these sub questions will provide a better understanding of the required changes in the processes and are the starting points of further interventions.
- What is the relative impact of patient, demand, and treatment characteristics on the outcome measures i.e. mortality, quality of life, and lead times?
- How do patient, demand, and treatment characteristics influence the planning of sur- gery and discharge?
- What is the relation between the planning of surgery and discharge and the outcomes measures of the treatment i.e. mortality, quality of life, and lead time?
- What are the actionable variables which can change the present processes in order to improve the planning and organization within the clinical pathway of patients of 60 years and older with a hip fracture?
1.4 Structure
The study starts with a description of the current processes with relation to the planning issues at the department of Orthopedics in chapter 2. Subsequently, in chapter 3 the definition of the problem is elaborated in a conceptual framework. The methodology of the study is stated in chapter 4 and in chapter 5 the results of the quantitative and qualitative analysis are listed.
Next, the discussion of the results is elaborated in chapter 6. Finally, the conclusions, recom- mendations, and the outline for further research is described in chapter 7.
2 SYSTEM DESCRIPTION
The study was performed on the department of Orthopedics and Traumatology of the UMCG.
In paragraph 2.1, the organization and the two departments are described. Subsequently, the routing of the patients during the treatment of the hip fracture is elaborated in paragraph 2.2.
Next, the different planning issues for the research population are described in paragraph 2.3.
2.1 Organization
The UMCG is the result of an alliance between the former Academic Hospital Groningen – AZG – and the medical faculty of the University of Groningen and exists in its current ap- pearance from 13th January 2005. The organization is one of the 120 hospitals and one of the 8 academic medical centers in the Netherlands. It provides work for more then 7000 people and has revenue of 719.0 million euro in 2007. Their mission is “Building the Future of Health”. This results in the following vision: Pioneering in research, examine and sharing knowledge, and taking care for people. These issues are listed in the core businesses of the hospital 3.
The activities of the medical specialty of Orthopedics involve the diagnosis, treatment, and rehabilitation of patients with a disorder of the musculoskeletal system. The department en- closes a nursing department with 31 beds, an outpatient clinic, and a number of researchers. A total of 70 people work at the department – among which 7 orthopedic surgeons and 9 assis- tant surgeons. On a yearly basis, around 5.600 patients are treated at the outpatient clinic, 1200 admissions take place, and – on day care alone – 600 patients get surgery 4.
The medical specialty of Traumatology is responsible for the care to patients who suffer from injuries caused by accidents. Since they provide acute care, this discipline is often associated with emergency medicine. The composition of the department is similar to that of Orthope- dics with a nursing department of 30 beds and a polyclinic. A total of 60 people are employed including 6 surgeons and 3 assistant surgeons 5.
2.2 Routing of the patients
In the analysis of the present performance, it proved important to determine the routing of the research population throughout the UMCG. This analysis leads to the overview in figure 2.1.
Figure 2.1 Routing of the patients during the treatment in the UMCG
The routing starts at the “Centrale Spoed Opvang” – CSO – where the diagnosis of the hip fracture is set. If the surgeons determine that surgery is required, the patient is transferred to the nursing department of Orthopedics or Traumatology. The patient is prepared for surgery and moved to the surgery centre. After the surgery, the patient returns to the nursing depart- ment of origin and is rehabilitated back to health. The choice for the location of discharge depends upon multiple factors among which the condition of the patient and the required care in the period after discharge are the most important ones. Often, the location of discharge is the pre fracture location – PFL – or a nursing home. The majority of the patients who are transferred towards a nursing home have a temporary stay until they can return to their PFL.
However, a number of patients already live in such housing or are unable to return their pre- vious homes due to their condition. A small percentage will decease during their stay in the hospital or will be transferred to other departments within the UMCG or other hospitals.
If the patient is treated without surgery, the staff determines whether returning to the PFL is possible. If so, the patient is discharged from the CSO and returns towards the PFL. If not, the patient is admitted at the nursing department before transfer to the nursing home.
The procedures with relation to the planning of the surgery and the discharge start when the patient is admitted at the nursing department. After the patient is approved for surgery – i.e.
all the medical tests are performed –, the planning employees apply for an operating room at the surgery center. After the social diagnosis is finished, an assumption is made about the probability that the patient can return to the PFL. If not, the patient is enrolled for a stay at a nursing home.
Arrival at CSO
Admission at nursing dep.
Surgery centre
Readmission at nursing dep.
Nursing home
Return to- wards PFL Sur-
gery
Transfer towards other dep.
Return to PFL
PFL Admission at
nursing dep.
Nursing home
No
No
Yes
Yes
Patient deceases
2.3 Planning issues
In paragraph 2.2, the process of the treatment was elaborated. Within this process, three plan- ning moments can be identified. These are:
o The admission planning of emergency patients;
o The surgery planning of elective and emergency patients;
o The discharge planning and transfer of patients towards nursing homes or the PFL.
In this paragraph, the issues of each planning moment are elaborated on the required re- sources, the priority rules and the current performance.
2.3.1 Admission issues
The available resources at the UMCG limit the number of admissions for patients with a hip fracture. Though a variety of departments and resources are involved during the treatment, just a few have substantial influence on the number of admissions. For the research popula- tion, the availability of sufficient hospital beds at the nursing departments of Orthopedics and Traumatology and the presence of qualified personnel to nurse the patients determine the number of admissions. These two resources are indicated as the bottleneck capacities.
In the present situation, a number of planning rules are used to regulate the flow of patients towards the UMCG. The hospital beds and personnel are the shared resources for patients with a variety of medical diagnosis for the two disciplines. Therefore, competition exists be- tween the different patient groups for admission. Within the UMCG, there is a policy of first come, first served.
According to Hanna and Sethuraman [2005], the capacity issues in the health care setting are not the result of the capacity level of the hospitals6. It is the unstable demand and the lack of adequate buffers. They propose, in line with Jack and Powers [2004], volume flexibility as the means to optimize the overall utilization and performance7. In the UMCG, a tactical under- utilization is pursued at the nursing departments to anticipate on the admission of emergency patients. However, there are no rigid regulations with relation to the height of the desired utilization and there are no beds dedicated to emergency patients. The desired utilization is dependent upon a number of factors. First of all, a portion of the capacity is used for planned patients. The exact number differs per week and is dependent upon the available operating rooms for planned surgery and the rehabilitation period in the hospital. Second, the utilization is determined by the desired service level of the hospital. The customer service level is the
ability to satisfy customers’ delivery data. In this setting, it is determined by the ability of the UMCG to admit emergency patients.
2.3.2 Surgery issues
Every surgeon of the department of Orthopedics is qualified to perform surgery on patients with a hip fracture. This enables the composition of the circulation scheme for the surgery of emergency patients. This scheme takes care that – at all times – a surgeon and at least one assistant surgeon of the department of Orthopedics are on duty to perform surgery on emer- gency patients. In some cases, the surgeon on duty postpones the surgery to a later date – re- gardless of the availability of other resources. This procedure is conducted since there are a number of surgeons specialized in this type of surgery. Based on the severity of the fracture and the additional diagnosis, the presence of such a specialized surgeon might be required.
The schedule of the operating rooms is determined by the planning staff of the surgery centre.
If an operating room is set apart for a patient with a hip fracture, this implies that the appro- priate number of anesthetists and surgery assistants are arranged. Next to the emergency oper- ating rooms, the departments of Orthopedics and Traumatology pursue a tactical under- utilization in the surgery scheme of the operating rooms for planned patients. It is necessary to reserve utilization in case planned surgeries last longer. Though this creates more capacity for emergency patients tool
The official policy of the surgery centre is to enable emergency surgery within the restrictions set in the USA-classification8. This classification indicates the maximum due date and time of the surgery and is issued to each patient at the CSO. Based on this classification, the patient should be treated immediately (code A of “acute”), within 6 hours (code S of “speed”), or within 24 hours (code U of “urgent”). Since most of the patients are classified as “Urgent” or are treated as one, surgery should take place within 24 hours. This measure is in line with the national regulations for hip fractures9. There are two priority rules with relation to the emer- gency operating rooms. First of all, an acute surgery is performed before a speed one and a speed surgery is performed before an urgent one. Second, per classification a rank is deter- mined based upon the earliest due date of the surgery.
Until now, the performance on the USA-classification and the maximal lead time before sur- gery is not used for further actions. Since there are no binding consequences associated to these measures, it lacked on necessity. The only rigid measure is the calendar day norm of the
“Inspectie voor de gezondheidszorg” or IGZ. This measure states that – dependent upon the ASA-classification of the patients – a certain percentage of the patients should receive surgery
at the calendar day after admission. The ASA-classification is a measure for the severity of the pre-existing morbidity and is derived from the organization which developed it - the American Society of Anesthesiologists10.
2.3.3 Discharge issues
The discharge of patients is possible after the medical necessity for further stay at the nursing department is gone. In the majority of the cases, elderly patients with a hip fracture are dis- charged to a nursing home or the pre fracture location. The required resources differ per loca- tion of discharge. If the patient is discharged to their pre-fracture location, additional home care and other support are necessary. This is arranged by the transfer nurse of the department.
The required level of care differs between patients due to the support of relatives or friends and is focused on a safe rehabilitation of the patient in a familiar environment. If the patient is discharged towards a nursing home, the registration for a placement and the supply of infor- mation is the responsibility of the transfer nurse. In general, the patients are discharged to- wards nursing home in and around the city of Groningen. Dependent on the available capacity of the nursing homes and the preferences of the patient, placement closer to the pre fracture location is possible.
From the point of view of the UMCG, there are no rigid priority rules with regard to the dis- charge of patients. External organizations on the other hand use the FIFO-principle to assign the home care or the available capacity at a nursing home. Most patients make use of this support for a short period of time. This aspect makes it relatively easy to arrange the required care, while long waiting times exist for long term treatment.
The current performance on the discharge procedures is difficult to determine. In 2008, the lead time of patients at the department of Traumatology was analyzed. This showed that a reduction is possible by eliminating the unnecessary stay at the department. The unnecessary stay existed especially in the lead time after surgery and involves an inefficient and delayed planning of the discharge towards nursing homes and arranging home care. A stronger focus on these issues has already improved the lead times of this department. Though the depart- ment of Orthopedics already performed better on the discharge procedures, a further im- provement on the lead time after surgery should be possible.
3 THEORETICAL FRAMEWORK
This chapter entails the model which will provide an answer on the research questions. First, the theoretical framework is presented in paragraph 3.1 in order to provide an overview of the relevant variables in this study. It starts with a description of the mutual relationship between the patient, demand, and treatment characteristics in paragraph 3.2. Subsequently, the three outcome measures are analyzed. Each paragraph contains the operationalization of the out- come measure, the prospective results, and the expected predictors based on the preliminary literature study and interviews. The analysis begins with the survival among patients with a hip fracture in paragraph 3.3. Next the measures related to the quality of life are elaborated in paragraph 3.4. Finally the different stages of the lead time are listed in paragraph 3.5.
3.1 Theoretical framework
In an attempt to provide an answer on the research questions, a model was derived from the preliminary investigation. First, a literature study on medical research of patients with a hip fracture was performed to identify the relevant variables and the expected relationships. This resulted in the theoretical framework in figure 3.1.
Error!
Figure 3.1 The theoretical framework of this study
Within this framework, a quantitative and a qualitative part can be identified. The quantitative part involves the expected relationship between the patient, demand, and treatment character- istics on each other. Next to this, it consists of the influence of these variables on the planning
4. Planning characteristics
1. Patient characteristics
3. Treatment characteristics 2. Demand
characteristics
5. Capacity 6. Criterions
7. Outcome measures
C
B A
D E
F G
H
characteristics and the outcome measures of this study – survival, quality of life, and lead time. The qualitative part entails the planning characteristics of admission, surgery and dis- charge for patients with a hip fracture and the corresponding influence of respectively the criterions and the capacity.
3.2 Patient, demand, and treatment characteristics
After the diagnosis is set at the CSO, the patient is transferred towards the nursing department of Orthopedics or Traumatology. By this time, the characteristics of the patient, the demand, and the desired treatment are known. The variables involved shape the first part of the theo- retical model, which altogether affects the planning mechanisms and the outcome measures.
The paragraph starts with an overview of the several variables in the three groups. Then, the excepted mutual relationships are described with the use of a literature study.
The patient characteristics of the research population are the age, gender and the pre fracture location of the patients. There are two assumptions with relation to the mutual cohesion of these characteristics. The first assumption is that female patients are of higher age due to the difference between life expectancy for men and women11. Next, different articles suggest that the majority of the patients is likely to be female 9, 12, 13, 14, 15
. The second assumption involves the pre fracture location, it is presumed that patients admitted from nursing homes are older then patients who lived independently or in a residential care centre.
The type and classification of the fracture, the part-time of admission at the CSO, the year of surgery, the ASA-classification, and the ASA-group are the demand characteristics for the re- search population. There is no obvious mutual cohesion between these characteristics except between the ASA-classification and the ASA-group. The cause of this cohesion is that the results on the characteristic ASA-group is derived from the results of the characteristic ASA- classification. In the medical research used, there are little differences between the prevalence of the two types of fracture – inter capsular versus extra capsular 14,15,16. The ASA- classification is an indication of the general condition of the patient – or the level of morbidity – and ranges from 1 till 5. The expectations are that the majority of the patients have an ASA- classification 2 or 3 12, 15, 16. This varies between diseases which have no effect on daily activi- ties till symptomatic diseases with little restrictions for the patient. The remainder variables – the classification of the fracture, the year of surgery, and the part-time of admission – are too specific for a reliable comparison with previous medical research.
The treatment characteristics are the department of admission, the type of implant, the implant choice, the type of anesthetics, the type of surgeon, the USA-classification, and the part-time of surgery. It was possible to determine two assumption from the interviews. The first as- sumption is the USA-classification which influences the part-time of surgery. Namely, the majority of the patients has an USA-classification “U” and these patients get – according to the policy – surgery during the office hours. This means that surgery takes place in the morn- ing and afternoon. The second assumption is the relation between implant and department of admission. A surgery which involves the placement of a total hip arthroplasty – THA – is al- ways performed by surgeons of the department of Orthopedics, which influences the choice for the department of admission. Furthermore, the results on two variables will show random- ized results. In the research period, the patients are admitted at the department of the surgeon on duty at the CSO and the type of anesthetics is often involves the preference of the patient and or the anesthetist. Next to this, the type and classification of the fracture determine the type of implant 9.
The patient, demand, and treatment characteristics include a total of 13 variables and these are listed in table 3.1.
Patient characteristics Age
Gender
Pre fracture location
Demand characteristics Type of fracture
Classification of fracture Part-time of arrival at the CSO Year of surgery
ASA-classification ASA-group
Treatment characteristics Department of admission Type of implant
Implant choice Type of anesthetics Type of surgeon USA-classification Part-time of surgery
Table 3.1 An overview of the variable of the patient, demand, and treatment characteristics
It is possible to identify a number of mutual correlations. The patient characteristics influence variables of the demand and treatment characteristics – relationship A and B. Both the age and the gender are predictors of the level of morbidity, which is measured with the ASA- classification 12, 17. Subsequently, the implant choice can be age-related. According to the in-
terviews, older patients might receive other implants, though the choice differs between sur- geons.
Based on the interviews, the demand characteristics have a huge influence on the treatment – relationship C. The type and classification of the fracture are an indication for the type of im- plant, though the surgeon makes the final decision 12. Next to this, the condition of the patient – measured in the ASA-classification – can influence the type of anesthetics. Lastly, the com- bination of type of fracture and implant choice determines the USA-classification 8.
3.3 Survival
The survival of patients is the most important outcome measure and therefore frequently in- vestigated in the medical research regarding hip fracture. The operationalization of this out- come measure is the mortality among patients within one year of surgery. First, the expected mortality rate among the research population is determined. Next, the expected relationships between mortality among the research population and the patient, demand, treatment, and planning characteristics are described. Then, the influence of the actionable variables is high- lighted.
Inevitable, a certain percentage of the patients will decease within one year after the fracture and this is – although regretful – impossible to influence 18. One of the causes is the average mortality rate for the Dutch population of 60 years and older. This was 10.09% in the research period. In addition to this, patients with a hip fracture have a higher risk of decease within one year of surgery. Though it lacks on data of the research period, the average of the previous years is a reliable indication of the expected mortality rate of the research population. In the period between 2001 and 2005, the mortality rate within one year after surgery for patients of 65 years and older in the Netherlands was 25.1% 19. However, these official numbers are – according to the Rijksinstituut voor Volksgezondheid en Milieu (RIVM) - subject to underre- porting. Due the lack of consistency in the registration of deaths, it is plausible that the actual mortality rate is higher. As described by Foss and Kehlet [2005], medical research on this subject shows divergent percentages 18. The percentages vary between 14% and 36% which makes it difficult to establish a reliable comparison 20.
A number of relationships between mortality and the patient, demand, and treatment charac- teristics can be identified. The starting point is the national regulation of Van Vugt et al [2007], which state a five some of variables which influence the mortality rate within one year of surgery 9.
The first variable in this and other studies is the age of the patient at the moment of fracture
9,12,16,21,22
. The saying “old age has its infirmities” is correct with regard to the patient popula- tion. Due to ageing, the average condition of humans decreases and makes the risk of mortal- ity higher – just like the elderly population as a whole.
The second variable of the study is the mobility of the patients which is often an indication of the overall condition. According to Berglund-Roden et al [1994] the general walking ability affects the mortality within 4 months and Hommel et al [2008] described the significant influ- ence of walking with a stick on mortality within 1 year 12,23.
Morbidity – the operationalizations in this study are the presence of pre-existing morbidity and complications due to the fracture – is regarded as the third factor which contributes to the probability of early decease. This statement is in line with research by Jacobsen et al. [1992], who determined that the level of morbidity influences the mortality rate 17. This effect was most significant at one and two year postoperative. Research by Claque et al [2002] proved the relation between level of morbidity and both in-hospital as ninety days mortality 15. In these articles, there was no distinction between pre-existing morbidity – measured as the ASA-classification – and complications due to the fracture.
The last two variables – gender and dementia – mentioned in the regulation support this statement. Though men are generally younger then women at the time of fracture, they suffer from more and / or severe morbidity 12. This might explain why mortality among men after 4 and 12 month is significantly higher. The same results were found by Holt et al [2008] who focused their investigation on early mortality 16. After 1 and 4 months, the percentages of de- ceased men were higher then for women, even when variables as morbidity were taken into account. Dementia on the other hand shows no direct link with mortality, though research by Dewey [1992] suggests that it does reduce the overall life expectancy of patients 24.
Reticence is necessary when the relationship between the pre fracture location and mortality is investigated. Crotty et al [2000] described the difference between patients who lived inde- pendent and those who made use of institutional care 25. Patients from nursing homes and residential care centers are – on average – older, have a higher risk of dementia, and a reduced mobility. Since all of these variables are mentioned above, it is likely that the effect of pre fracture location is the sum of other variables.
The relationship between the lead time before surgery and the mortality rate among the popu- lation is arbitrary too. There are various articles which support or challenge the mutual inter- dependency and it proves difficult to determine the cause effect relation. On one hand, the condition of patient might deteriorate while waiting on surgery 26,27. On the other hand, post-
ponement of surgery might be necessary to stabilize the condition of the patient 28,29. From a retrospective point of view, establishing a solid conclusion about the research population will be impossible. However, all authors agree that surgery for medically fit patients should take place as soon as possible.
The reduction of complications due to the fracture and the lead time before surgery are the two actionable variables with relation to the mortality rate. A study by Hommel et al [2008]
shows that patients admitted at an Orthopedics ward have lesser complications then those ad- mitted at other departments 30. The cause of this difference is assumed to be the experience of the nursing staff with this specific diagnosis. Since a number of medical examinations should be performed before surgery can take place, the lead time before surgery can only be influ- enced to a certain extent. Subsequently, the nature of the surgery – emergency surgery – makes it difficult to allocate resources in advance for this specific diagnosis. Finally, the gen- eral condition of the patient might require a delay of the surgery.
3.4 Quality of life
Due to the improved performance of the medical science and the increased expectations of the population, the importance of quality of life for patients made a progression in the last decen- nium. Within this study, two operationalizations are used to describe the performance on this outcome measure. These are the presence of complications due to the fracture and the ap- pearance of secondary surgery, which are described below
3.4.1 Complications due to the fracture
The presence of morbidity is associated with a substantial reduction of the quality of life for the patient and requires additional care by the nursing staff. A distinction can be made be- tween pre-existing morbidity and complications due to the fracture. The first group involves a wide variety of medical diagnosis, which deteriorates the condition of the patient and influ- ence the lead time before and after surgery. Though the UMCG is unable to influence the presence of this type of morbidity, it has a substantial influence of the outcome measure “lead time”. The second group involves complications which occur as a result of the fracture or the surgery itself. Van Vugt et al [2007] identified 8 common complications for patients with a hip fracture, from which a five some can be measured in this study 9. These are pneumonia, delirium, decubitus, wound infection, and urinary tract infection. Next to this, the appearance
of luxation is investigated. This type of complications has a huge influence on the rehabilita- tion and increases the lead time after surgery substantially.
The IGZ uses a number of complications as a performance measure for hospitals and one of these measures applies for the research population. This is the presence of decubitus or bed sores. The prevalence of decubitus in the Netherlands was 5.3% in 2006 and 3.8% in 2007.
However, this percentage is generally higher in the 8 UMC’s. The IGZ undertakes action when the prevalence is higher then 9.0% and 6.4% in respectively 2006 and 2007 31,32. The limitations of these figures are the methods of registration. Hospitals use different methods to measure decubitus and classification of decubitus lacks consistency.
In the majority of the medical research, there is no distinction between the two types of mor- bidity. Patients with a weaker condition are more susceptible to complications due to the frac- ture. Therefore, the presence of complications due to the fracture is associated with a higher age at time of admission and the presence of pre-existing morbidity. With the course of years, the condition of the patients deteriorate. This shows in weaker muscles and a more fragile condition. As a result, older patients are more susceptible for complications 27. Next to this, the presence of pre-existing morbidity is an predictor for the presence of complications due to the fracture. This presence is an indication of a weaker condition, which results in a higher risk 22. Finally, the presence of contra indication for early surgery increases the risk of certain complications. Therefore, failing to meet the calendar day measure could be a predictor of complications due to the fracture.
3.4.2 Secondary surgery
As a result of complications as mentioned in paragraph 3.4.2 or problems with the implant, it might be necessary to perform secondary surgery on a patient. A case of secondary surgery is defined as any surgery within a timeframe of one year after the initial procedure which is re- lated to the fracture. It is performed when the initial surgery did not lead to the desired result or when severe complications arise from it. The prevention of this type of surgery will sub- stantially increase the quality of life and the effectiveness of the departments. However, there is no legislation with regard to the acceptable prevalence of secondary surgery. The type and nature of these surgeries make it difficult to establish a consistent policy and judgment for this kind of surgeries.
There are several points of interest with regard to secondary surgery. The first issue is deter- mining whether the cause of the surgery is the result of the performance at the primary sur- gery. Specific types of secondary surgery – like removing screws to reduce pain – and surgery
within a short time frame after the initial one can be related to the initial procedure. However, other issues arise in the course of time. Therefore, the choice for the time frame is set on one year after surgery. The second issue is that a distinction is necessary between planned and unplanned surgery. The impact of planned surgery is substantial lower then unplanned sur- gery given the reason of surgery. The third and final issue is that secondary surgery is sub- stantially dependent upon the type of fracture and implants used 12. However, it is possible to identify a number of general causes.
A number of articles mention this phenomenon, though these show that the average frequency is hard to determine. A Danish research by Foss et al [2007] found a percentage of 19%
among 600 patients within a period of six months after surgery 14. Subsequently, an investiga- tion over a period of two years by Mjørud et al. [2006] among 199 patients showed a rate of 25.1% 33. This could mean that most cases of secondary surgery appear within a short period after the initial one, which is in line with the nature of complications. On the other hand, Hommel et al. [2008] performed a research among 420 patients and determined that 9% of their population underwent secondary surgery within one year after the initial one 12. These conflicting outcomes underline the heterogenic character of the research population and the implications on the outcomes. Other causes of the different outcomes are the size of the re- search population – a small population increases the risk of a distorted image of reality – and the research period used.
The appearance of secondary surgery can be linked to the specifications of the implant and the presence of certain complications due to the fracture. According to Palm et al. [2007] sur- geons with more experience have a lower appearance with secondary surgery among fractures with a high classification. Since the UMCG is an academic hospital, more surgeries are per- formed by assistant surgeons 40. Furthermore, patients who receive planned secondary surgery are – on average – younger. Those patients have a better condition which enables this type of surgery 12.
3.5 Lead time
The majority of the research on hip fractures is focused on the outcome measures survival and quality of life. In recent years, the attention shifted more towards the lead time of the treat- ment due to the increasing pressure of cost reduction in health care settings. Since most au- thors limit their research at the lead time before surgery and the total lead time, it lacks on reliable predictors for the lead time at the surgery center. The lead time after surgery shows a significant cohesion with the total lead time and these are therefore discussed simultaneously.
This paragraph starts with a description of the necessity to reduce the lead time at the hospital.
Next, the lead time of the UMCG is compared with other UMC’s and top clinical hospitals to determine the present performance. Then, the remaining of the paragraph describes the pre- dictors of the three individual stages of the lead time.
3.5.1 The need for lead time reduction
The lead time at the UMCG is measured in the number of days the patient spends in a hospital bed. According to the “Diagnose Behandeling Combinatie” – DBC – for hip fractures, the UMCG receives 1050 euro for the day of admission and 50 euro for the remainder days the patient spends in the hospital. However, the costs for a hospital bed at the department of Or- thopedics or Traumatology are set at 300 euro by the Board of Directors of the UMCG. This means that after 4 days, the income and expenses cancel each other and that the remainder of the lead time should be compensated with other means. Obviously, the reduction of the total lead time will significantly improve the cost effectiveness of the departments. Yet the meth- ods which lead to the best result are unknown.
In order to assess the present performance in terms of lead time, it is analyzed in comparison with those of other suppliers. Due to the international differences in the organization of care, it proves difficult to compare research results between the UMCG and academic literature.
Therefore, the performance of the UMCG will only be compared with other UMC’s. These organizations treat the same group of patients which results in a reliable indication of the per- formance. In 2007, the average lead time for all UMC’s was 14.1 days versus 12.6 days at the UMCG. Only the Leids University Medical Center at Leiden performed better on this out- come measure. On the other hand, general hospitals – among which the CWZ in Nijmegen and the Twenteborg in Almelo – are able to discharge the patients within one week. Though the patient population between an UMC and these hospitals might differ, an improvement of the performance by the UMCG should be possible.
The lead time in the UMCG is composed of three stages which together form the total lead time. A visualization of these stages is made in figure 3.2.
Figure 3.2 Composition of the lead time of the treatment.
Lead time before surgery
Lead time at surgery center
Lead time after surgery
Total lead time of the treatment
3.5.2 Lead time before surgery
The lead time before surgery is the only stage of the lead time which is judged by the IGZ with relation to the performance of the hospital. According to van Vugt et al [2007], the pa- tients should receive surgery within 24 hours after admission in the hospital unless contra indications exist 9. Next to these medical reasons, the causes for a prolonged lead time before surgery can be capacity issues and administrative procedures with relation to the planning of the surgery 12. According to the literature, the predictors for the lead time before surgery are the USA-classification of the patient, and the type of fracture 8,12. In addition to this, the ASA- classification came up for discussion during the interview as a cause for an increased lead time before surgery.
The USA-classification is a measure used in the UMCG to indicate the urgency of the care needed. Therefore, a higher classification – A is the highest and U is the lowest – implies a shorter lead time before surgery 8.
The ASA-classification is an indication of the general health of the patients 10. A higher classi- fication indicates a poorer condition and this might result in the need for consultation and / or a postponement in order to stabilize the patients. On the other hand, these patients might re- ceive early surgery in order to prevent their condition from getting even worse. Given the retrospect character of this study, it proved impossible to evaluate these data unambiguously.
Therefore, it is possible that this variable has an insignificant influence on the lead time be- fore surgery in this study.
The last predictor according to the literature research is the type of fracture. According to Hommel et al [2008], extra capsular fractures are associated with a shorter lead time before surgery 12. One of the explanations is the nature of the fracture. In case of a hip fracture, there is always the risk of slow death of the hip due to the lack of blood supply. This slow death is a lesser problem when the hip is replaced by a prosthesis, yet extra capsular fractures are treated with screws. Therefore, it is more likely that an extracapsular fracture has a shorter lead time before surgery. Since Hommel et al [2008] did not perform a multivariate analysis on their results and other research shows little support on this subject, its influence is arbitrary.
3.5.3 Lead time at surgery centre
The lead time at the surgery centre is just a small part of the total lead time. Next to this, its influence on the other stages of the lead time lack in the national regulations of van Vugt et al.
and academic research 9,12. Therefore, the predictors of this stage of the lead time are specula- tive.
First of all, it is likely that a more experienced team performs the surgery in a short period of time. In the research period, the definition of experience is set on the type of surgeon and the department of admission. It is likely that an assistant surgeon has lesser experience with per- forming surgery on a hip fracture and this might result in a prolonged lead time at the surgery.
Furthermore, the surgeons at the department of Orthopedics are more familiar with the place- ment of a prosthesis – due to the planned surgeries – which suggest that this department has a shorter lead time on this type of surgeries.
Second, the actions taken by the surgeon differ dependent on the implant choice. Placement of a prosthesis is more thorough then placing 3 screws, which results in differences in lead times at the surgery center.
3.5.4 Lead time after surgery and the total lead time
The predictors for the lead time after surgery show huge similarities with those of the total lead time. This resemblance is the result of the overlap between the duration of both lead times. Since early surgery – within 24 hours after admission – is suggested for the research population, the majority of the total lead time consists of the lead time after surgery. Based on the national regulation of Van Vugt et al [2007] and other research, it is possible to define morbidity, gender, pre fracture location of the patient, the department of admission, and the lead time before surgery as the predictors of the lead time after surgery and the total lead time
9,12,15,16,30
. Subsequently, the predictors of the total lead time after surgery can be derived from those of the lead time after surgery.
The first and most important predictor of the lead time after surgery is the presence of mor- bidity – both pre-existing morbidity and complications related to the fracture. These delay the recovery of the patient and cause the necessity for additional care at the nursing department 15. The influence of gender on the lead time in a hospital is an arbitrary one, given the outcomes on multiple studies 12,15,16,30
. Hommel et al [2008] showed that male patients have a shorter total lead time 12. It is assumed that male patient have more often a partner at home to provide support after the fracture. This is due to the difference in life expectancy between men and women. Subsequently, the discharge towards the pre fracture location should be less compli- cated for male patients. However, the argument is not applicable if the patient is discharged to a nursing home or if home care is arranged and available on time.
The relationship between pre fracture location and the lead time after surgery is less arbitrary.
Since patients require support after discharge from the hospital, – additional – home care is required or even temporary placement in a nursing home. These procedures are redundant if
the patient is admitted from a nursing home in the first place. Next to this, the doctors at these facilities are qualified to monitor the rehabilitation of the patient, which enables an earlier discharge too. Therefore, a distinction is made between patients from independent living and residential care centers versus patients from nursing homes.
According to the national regulations by Van Vugt et al [2007], a patient receives an optimal treatment at a department which is familiar with this specific diagnosis 9. Therefore, a patient with a hip fracture should be treated at the department of Orthopedics or Traumatology. Due to the limited capacity at these departments, patients in the UMCG can be admitted at other departments which results in prolonged lead times after surgery. Next to this, it is expected that patients of the department of Traumatology have a longer lead time after surgery. This is due to the inefficient use of the discharge procedure in the research period.
Finally, the last predictor for the lead time after surgery is the lead time before surgery. The stage of the lead time is a popular subject in medical research, due to its suspected yet arbi- trary influence on the lead time after surgery. Some researchers argue that a prolonged lead time before surgery leads to more and / or severe complications due to the fracture which af- fect the lead time after surgery 9. Others pretend that the presence of morbidity is a contra indication of early surgery in the first place, which means that these patients have a poorer health. Subsequently, these patients spend more time in the hospital. As a result, a general rule is established that patients should receive surgery within 24 hours after admission in the hospital unless contra indications exist.
3.6 Admission planning
The admission of emergency patients with a specific diagnosis – age of 60 years and older with a hip fracture – is hard to predict. Jack and Powers [2004] developed a volume flexible strategies framework, which present four optional strategies for each combination of the de- mand uncertainty versus the range of internal flexibility 7. The research population of this study are emergency patients with a specific diagnosis. Therefore, the demand uncertainty with relation to these patients is high. The range of internal flexibility is high too. These pa- tients can be admitted at multiple departments in the UMCG, when Orthopedics and Trauma- tology lack available capacity. In addition to this, planned surgery on patients for these two departments can be postponed when emergency patients require the hospital beds at the nurs- ing departments. Therefore, the appropriate strategy for the research population is the mitigat- ing strategy.
The mitigating strategy enables the UMCG to adapt a high level of volume flexibility. There are three common approaches which can be used independently or in combination with the other two approaches:
o Restructuring and reallocation of facilities; This involves multiple techniques among oth- ers the resource sharing. This is applicable when demand is high – referring patients to other departments – or low – reduce the number of hospital beds.
o Risk pooling; With the use of risk pooling strategies, the demand for the different services can be aggregated. Given the shared resources – hospital beds and nursing staff for a vari- ety of diagnosis – of the departments, the capacity is used effectively.
o Outsourcing and strategic alliances; The choice for outsourcing can be found in cost re- ductions, improved service level, and gaining a higher market share. External networks or strategic alliances can be used to control the source of the demand or supply and improve the synchronization between the stages of the supply chain. With relation to the research population, this involves the cooperation with nursing homes and – to a lesser extent – home care facilitators.
4 METHODOLOGY
This chapter entails the methodology of this study and the corresponding considerations made. It starts with the research strategy and the realization of the patient population. This is followed by the method of data collection and analysis, and the operationalizations of the sub- research questions.
4.1 Research strategy
The research strategy describes and explains events in reality, which makes it part of the em- pirical science. Furthermore, this research pursues the acquisition of knowledge concerning the processes of planning and organization of the department of Orthopedics and Traumatol- ogy. Due to these characteristics, the research design is based on a literature study in combi- nation with an exploratory case study. It encloses a retrospective analysis of the patients of age 60 years and older who were admitted with a hip fracture in the years 2006 and 2007.
With this analysis, the different variables and their influence on the lead time and other out- come measures were investigated.
4.2 Research population
The number of cases of hip fractures was determined with the use of the DBC-code. This code stands for the combination of the diagnosis and the treatment, which is used for the fi- nancial settlement of the treatment. In the research period – the years 2006 and 2007 –, a total of 257 cases were registered in the UMCG. Since the focus of this study is on elderly patients – age of 60 years and older –, 95 cases were removed from the population. Then, 15 cases were excluded based on deviant implants, in order to perform a reliable and consistent analy- sis on a more homogeneous population. Finally, one additional case was excluded since the patient received implants on both hips during one surgery. Since one patient was admitted twice for a hip fracture – fracture on the left and the right side -, this leaves 145 cases of 144 individual patients in the research population. It involves 83 inter capsular fractures and 62 extra capsular ones.
4.3 Data collection and variables
The rate of mortality and the quality of life is influenced by the factors mentioned in the con- ceptual framework. A literature study of academic papers provided insight in the relevant is- sues concerning the outcomes measures of surgery among patients with hip fractures. The
meetings with the steering committee and interviews with key players – planning personnel, surgeons, and head of nursing department – provided the necessary information with regard to describing the current processes in the UMCG. Since the hospital use different systems to archive the patient related data, it was hard to determine the exact size of the patient popula- tion. Eventually, the database was composed of data from three different systems and a previ- ous study by an anesthetist. The missing data was gathered by an analysis of the patient files.
In addition to this, an extensive verification took place to ensure the quality of the data.
The conceptual framework was described in the previous chapter and a total of 7 parts were identified. These have a number of variables which are listed below. In parentheses, the op- erationalizations of these variables were described.
1 Patient variables:
o Age;
o Gender (male versus female patients);
o Pre fracture location (living independent or in a residential care center versus living in a nursing home).
2 Demand variables:
o Type of fracture (intercapsular versus extracapsular);
o Classification of fracture (the severity of the fracture, ranging from Garden I till IV);
o Year of surgery;
o Morbidity (regardless of the nature);
o ASA-classification (measure to indicate the level of pre-existing morbidity, ranging from 1 till 5);
o ASA-group (ASA- classification 1-2 versus ASA-classification 3-5);
o Complications due to the fracture.
3 Treatment variables:
o Department of admission (Orthopedics versus Traumatology);
o Type of implant (screws versus prosthesis);
o Implant choice
3 screws
DHS
KHA
PFN
THA
o Type of anesthetics (general versus regional);
o Type of surgeon (assistant surgeon versus surgeon);
o USA-classification (priority measure for patients who receive emergency surgery);
4 Criterions with relation to the treatment o Calendar day norm by the IGZ;
o USA norm by the UMCG.
5 Planning:
o Planning of admission;
o Planning of surgery;
o Planning of discharge location.
6 Capacity variables:
o Beds
Nursing department;
Operating rooms;
Nursing home.
o Personnel
Surgeons;
Anesthetists and surgery assistants;
Nurses at the departments of Orthopedics and Traumatology.
o Part-time of surgery (surgery by day – between 6.00 A.M. and 6.00 P.M – versus surgery by night – between 6.00 P.M. and 6.00 A.M).
7 Outcome measures:
o Survival rate among the population
Mortality within one year after surgery (irrespective of the cause of death).
o Quality of life
Complications due to the fracture during the lead time in the hospital;
Secondary surgery within one year after surgery.
o Lead time in the hospital
Lead time at CSO;
Lead time before surgery;
Lead time at the surgery center;
Lead time of the surgical procedure;
Lead time after surgery until discharge;
Total lead time.
Most of the data was provided by a consultant of sector B and gathered from the electronic patient files. The missing data was collected with the use of two systems – POLI-plus and OK-plus – which contained the same data as the patient files. The data with relation to the time and day of arrival at the CSO was provided by secretary of the CSO and the objectives were defined in the regulations with relation to the desired treatment for patients with a hip fracture 9. The interviews with the project leader, surgeons and the planning employees pro- vided the information about the capacity of the departments and the description of the proc- esses. Finally, a number of variables were derived from the information gathered, such as the different stages of the lead time that were calculated with the use of the admission, surgery, and discharge times.
4.4 Data analysis
The data analysis consists of both a quantitative and a qualitative part. This paragraph starts with the quantitative part, which entails the patient, demand, and treatment characteristics and their influence on the outcome measures. Next, the qualitative part describes the planning issues with regard to the admission, surgery, and discharge of patients.
For the quantitative part of the data analysis, a database with 145 cases of hip fractures is used. This database was established in Microsoft Office Excel 2003 and used the data from a number of computer systems. Then, the database was imported in SPSS version 16.1 to per- form the analysis on the outcome measures. Before this analysis took place, the outliers for each part of the lead time were identified and removed from the population. Most articles on this subject included outliers in their study, since their interest lay in presenting a realistic view of reality. This study focuses on the average patient, which requires the removal of out- liers. An outlier was defined as any case with a value above or under the boundaries of the 95% confidence interval. Next to this, a distinction between the fractures types was made for the stages of the lead time. Given the differences in characteristics, an analysis of the total
