Master thesis in Pharmacy, 30 hp Master of Science in Pharmacy, 300 hp
Report approved: Spring term 2020
Supervisors: Magnus Ekström, Maria Sjölander, Examinor: Anders Öhman
Prescription of opioids and benzodiazepines for
breathlessness in
interstitial lung disease
A national population-based study
Jenny Genberg
Abstract
Introduction: Interstitial lung disease (ILD) is a large group of disorders that cause fibrosis of the alveoli, distal airways, and septal interstitium of the lungs, which in turn causes stiffness in the lungs and difficulty breathing. The clinical picture of ILD is dominated by breathlessness and worsening fitness. Breathlessness reflects the imbalance between the need to breathe (or ventilatory drive) and the ability to breathe (or ventilatory capacity). The term chronic breathlessness means disabling breathlessness that persists despite optimal treatment of underlying diseases/causes. Chronic breathlessness is present by almost all patients with severe ILD. The pharmacological symptomatic treatment for breathlessness, with the best evidence base, is systemic (non- inhaled) low-dose opioids. Despite evidence supporting the use of opioids for breathlessness, they are rarely prescribed for that indication in oxygen-dependent COPD.
Benzodiazepines may also be prescribed for breathlessness despite their lack of evidence for a beneficial effect for breathlessness. Due to the questionable effect of benzodiazepines for relieving breathlessness, and the fact that it is still prescribed for this indication, it is interesting to analyze the extent of it´s prescribing in ILD. No study has evaluated the indications for actual prescribed opioids and benzodiazepines for patients in ILD.
Objective: The aim of this study is to evaluate the stated indications and the medication types of dispensed opioid and benzodiazepine prescriptions for patients with severe ILD treated with long-term oxygen therapy in Sweden.
Method: This was an observational, population-based, longitudinal cohort study of patients with ILD aged ≥18 years, starting long-term oxygen therapy (LTOT) between October 1, 2005 and December 31, 2014 in the National Registry for Respiratory Failure (Swedevox). Random samples (n=2,000 for each treatment) of the participants’
dispensed benzodiazepine and opioid prescriptions during the study period were analyzed. The indications were examined over time during the observation period. All prescriptions from 6 months before LTOT (start of the study) until follow-up (first of death or study end) were included. The stated indications were categorized by type.
Results and Discussion: Breathlessness as a stated indication for benzodiazepines and opioids is uncommon in patients with severe oxygen-dependent ILD, also in the last months of life. This finding suggests that many of these patients are undertreated for their chronic breathlessness. The finding that opioids are prescribed more often for breathlessness than benzodiazepines, contradicts the fears that the prescription of benzodiazepines was greater than opioids despite their lack of efficacy and side effects.
The fact remains that benzodiazepines are still prescribed for this indication in some patients despite lack of evidence. The most predominant indication throughout the study period was pain for opioid prescriptions and anxiety for benzodiazepine prescriptions, which was expected since anxiety is one of the main indications for benzodiazepines as pain is for opioids.
Conclusion: In severe oxygen-dependent ILD, benzodiazepines and opioids are prescribed in about 40 percent of the cohort, but they are rarely prescribed for breathlessness, even in the last months of life. Which indicates that these patients are undertreated for their breathlessness and are at greater risk of suffering. The most common stated indication was anxiety for benzodiazepines and pain for opioids and the most common prescribed substances were Oxazepame and Oxycodone.
Key words: ILD, benzodiazepines, opioids, breathlessness, LTOT
Table of contents
1 Introduction ... 1
1.1 Interstitial lung disease ... 1
1.2 Breathlessness ... 1
1.3 Benzodiazepines ... 4
1.4 Opioids ... 5
2 Objective ... 7
3 Method ... 8
3.1 Design ... 8
3.2 Types of benzodiazepines and opioids ... 8
3.3 Method ... 8
3.4 Statistical analyses ... 9
3.5 Ethical considerations ... 9
4 Results ... 10
4.1 Baseline characteristics ... 10
4.2 Indications ... 12
4.3 Types of benzodiazepines and opioids ... 15
5 Discussion ... 16
5.1 Choice of design ... 16
5.2 Main findings ... 17
5.3 What this study adds ... 19
5.4 Strengths and limitations ... 20
5.5 Implications ... 20
6 Conclusion ... 21
7 Acknowledgement ... 22
8 References ... 23
1
1 Introduction
1.1 Interstitial lung disease
Interstitial lung disease (ILD) is a large group of disorders that cause fibrosis (scarring) of the alveoli, distal airways, and septal interstitium of the lungs, which in turn causes stiffness in the lungs and difficulty breathing [1]. ILD include diseases such as: interstitial pneumonia (infection of the interstitium by bacteria, viruses, or fungi), idiopathic pulmonary fibrosis (chronic, progressive form of fibrosis), nonspecific interstitial pneumonitis (autoimmune conditions ex. Rheumatoid arthritis), hypersensitivity pneumonitis (caused by long-term inhalation of dust, mold, or other irritants), cryptogenic organizing pneumonia (no infection), acute interstitial pneumonitis (fatal), desquamative interstitial pneumonitis (often caused by smoking), and sarcoidosis [2].
ILD is associated with poor survival and had a 52% increase in mortality between 2005- 2015, the cause of the increased mortality rate is unclear [3]. Higginson et al. analyzed all deaths with ILD as the cause of death between 2001 and 2014 [4]. The result was a total of 45,712 deaths, with a 9.2% increase each year, and 70.1% of these deaths occurred in hospitals. There was a peak of the hospital deaths between 2003-2005, but after 2005 the numbers reduced over time with 3-6%, probably due to the impact of a national end of life care strategy. The causes of ILDs are unknown in most cases but may include exposure to some chemicals, medications (methotrexate, leflunomide, amiodarone, nitrofurantoin and ethambutol) and medical interventions [1, 5]. Some autoimmune diseases can cause lung damage, for example rheumatoid arthritis, scleroderma, sarcoidosis and Sjogren´s syndrome [5]. Although the etiology is unknown, risk factors have been found and they are older age, male sex, cigarette smoking, hepatitis C, history of tuberculosis, history of pneumonia, older age, and chronic obstructive pulmonary disease (COPD) [6]. Many forms of ILD are irreversible and progressive but can be slowed down by medications [7].
The incidence of ILD is higher for men [8].
The most common form of ILD with the worst prognosis is idiopathic pulmonary fibrosis (IPF) [9]. Its etiology is unknown and cannot be reversed or cured, but there are two antifibrotic drugs that can delay the deterioration. The median survival time is around 3 years after given diagnosis, which is worse than many malignancies. IPF has often symptoms such as breathlessness and dry cough in the beginning and develops a restrictive lung function impairment over time [10]. Feeling sick, fatigue and weight loss may also be present, but fever, joint pain, and muscle pain indicates another cause (not IPF) [11]. IPF most often affects men aged 40-70 years [10].
The evaluation of suspected ILD firstly involves a careful review of the clinical history with focus on occupational history, environmental exposures, past and present drugs, and search for indications of systemic disease that might affect the lungs in a negative way [1].
Treatment of ILD varies depending on the underlying condition and patient factors [12].
It includes both treatment of underlying diseases and symptom relief [13]. The aim of the treatment is to preserve lung function, as the scarring and damage of the lungs is irreversible [14]. The treatment is mainly antifibrotic therapy that can slow the progression of the lung function decline but have not been shown to improve health- related quality of life or breathlessness. The only treatment that have prolonged life for selected patients is lung transplantation [10].
1.2 Breathlessness
The clinical picture of ILD is dominated by breathlessness and worsening fitness [1, 13].
Breathlessness is a subjective feeling of breathing problems and can´t be predicted by using only physiological parameters such as respiratory rate and common lung function measures [14]. Breathlessness reflects the imbalance between the need to breathe (or
2
ventilatory drive) and the ability to breathe (or ventilatory capacity) [15, 16]. It comprises several aspects (dimensions) that can be separated by the subject: the intensity of the feeling, degree of discomfort, the characteristics of the feeling (sensory quality), feeling that arise as a result (depression, anxiety, fear), and how it affects the function of the individual [13]. Breathlessness has a negative influence on life quality and may cause impaired activity and suffering usually in elderly people and in people with advanced disease, increased anxiety and depression, and increased risk of hospital admission and premature death [17-20]. The term chronic breathlessness means disabling breathlessness that persists despite optimal treatment of underlying diseases/causes [13].Chronic breathlessness is present by almost all patients with severe ILD [21-24].
Breathlessness is one of the most common causes of emergency hospital admission and visit, and leads to high health, social and informal costs [25]. Patients with terminal ILD may experience breathlessness more frequently than patients in end-stage lung cancer [26]. A study in Sweden compared the prevalence of symptoms and palliative treatments in the last week of life between patients with oxygen-dependent ILD and patients with lung cancer [27]. Patients with ILD suffered more from breathlessness (75%) than the patients with lung cancer (42%) (p<0,005), as shown in Figure 1. In the same study, the patients with ILD had also more unrelieved breathlessness, pain, and anxiety (Figure 2).
Figure 1. Prevalence, expressed as a percentage, of symptoms in interstitial lung disease (ILD) and lung cancer the last week of life in Sweden. The symptoms are breathlessness, death rattle (terminal respiratory secretions), pain, anxiety, nausea, and confusion. Reproduced from [End-of-life care in oxygen-dependent ILD compared with lung cancer: a national population-based study, Ahmadi Z, Wysham NG, Lundström S, Janson C, Currow DC, Ekström M, 71(6), 510-6, 2020] with permission from BMJ Publishing Group Ltd.
3
Figure 2. Symptom management with prevalence, expressed as a percentage, of relieved, partially relieved, and unrelieved symptoms in patients with interstitial lung disease (ILD) and lung cancer in the last week of life in Sweden. Reproduced from [End- of-life care in oxygen-dependent ILD compared with lung cancer: a national population-based study, Ahmadi Z, Wysham NG, Lundström S, Janson C, Currow DC, Ekström M, 71(6), 510-6, 2020] with permission from BMJ Publishing Group Ltd.Symptomatic treatment of breathlessness includes both pharmacological and non- pharmacological treatments, which can be seen in Figure 3 [13]. Non-pharmacological treatment can be psychosocial support, walking aids, and a handheld fan. Pulmonary rehabilitation has evidence, mostly regarding COPD but even evidence in case of ILD and cancer, that it decreases breathlessness, increases exercise capacity, and improves quality of life. It is classified as a safe and beneficial in patients with stable comorbidity (including respiratory failure). Multidisciplinary rehab includes endurance and strength training, improved breathing techniques, smoking cessation, and nutritional and psychosocial support which can be offered to patients. The pharmacological symptomatic treatment with the best evidence base is systemic (non-inhaled) low-dose opioids. From the evidence available, the initial dose is suggested to be equivalent to 10 mg oral morphine daily, and if needed titrated once a week until beneficial effect while balancing potential adverse effects. Long-term treatment with oxygen is indicated for patients with chronic lung failure, characterized by chronic severe hypoxemia which may be from fibrosis of the lungs [28]. Supplemental oxygen has inconsistent evidence for relieving chronic breathlessness and is mainly indicated for improving the oxygenation of the internal organs to prolong life.
4
Figure 3. Schematic picture of the principles of breathlessness management. The treatment is following this treatment staircase depending on symptom severity, availability, and cost of treatments, and in aspect of the beneficial and adverse effects for individual patients. Derived from reference [13].1.3 Benzodiazepines
Benzodiazepines are used mainly for anxiety disorders, acute epilepsy, alcohol withdrawal, insomnia and anesthetic premedication [29]. But they are sometimes also prescribed for breathlessness due to their sedating effect and therefore might have palliating effects for breathlessness. Benzodiazepines have no evidence for a beneficial effect for breathlessness or statistically significant effect in the prevention of episodic breathlessness. Use of benzodiazepines in patients with COPD may increase the risk of respiratory exacerbations and mortality and may increase the risk of asthma exacerbations [30]. There is no association between benzodiazepines and increased hospital admission, but a study found an association for treatment with high-dose benzodiazepines and that they may increase mortality [31]. Their effects are mainly sedative-, hypnotic-, muscle-relaxant, anxiolytic, and anticonvulsant [29, 32]. They have many side effects such as drowsiness, lethargy, fatigue, and at higher doses the following can occur; impaired motor coordination, dizziness, vertigo, slurred speech, blurry vision, mood swings, and euphoria [33]. Sever adverse effects can occur when administered together with other drugs such as opioids, due to additive or synergistic effects on the central nervous system and respiratory function. Respiratory depressant effects are enhanced greatly when co-administered with opioids but can also be a dose-related side effect. But according to a study, benzodiazepines are not often associated with adverse effects [34]. In a survey where hospice physicians and nurses in the U.S. were to indicate their level of agreement regarding benefit from benzodiazepines for end-of-life symptoms, the findings were that they found them beneficial [35]. And 39 % of the physicians thought that benzodiazepines are overused at their own hospice, although there is little supportive clinical evidence and a great potential for negative effects.
The mechanism of action of benzodiazepines is their interaction with specific subunits of GABA (gamma aminobutyric acid) receptors [29]. It is the ligand-gated chloride-selective ion channel, GABAA receptor, which is the target for benzodiazepines [33, 35]. By reading the information available on use of benzodiazepines for breathlessness, including studies, there is really no evidence for or against treatment with benzodiazepines for the relief of breathlessness as it is for opioids.
5 1.4 Opioids
Opioids are a widely studied agent in the treatment for breathlessness and reduces breathlessness in patients with a variety of conditions including COPD, ILD, cancer and chronic heart failure [16]. The evidence of treatment is only for short-term administrations, whereas evidence of long-term treatment is limited. They are often given as pain relief and their receptors are included in the family of more than 1,000 G-protein coupled receptors (GPCRs) [36-38]. When GCPRs are activated by agonists, an intracellular signaling cascade occur which mediates the actions of hormones and neurotransmitters, they decrease calcium currents within the cell bodies, which inhibits neuronal release and transmitter release [39]. There are four classes of opioid receptors which are the MOP (μ), KOP (κ), DOP (δ), and the nociception/orphanin FQ peptide receptor (NOP) [37]. Three of them have been localized in the respiratory tract [39]. The endogenous opioid system affects physiological effects such as pain, respiratory control, stress response, and appetite. MOP, DOP, and KOP also mediate exogenous opioids such as morphine and codeine. The lungs may also contain opioid receptors, for example pulmonary opioid receptors who are associated with vagal afferent C-fibers and the juxta- pulmonary capillary receptors (J-receptors) in the alveolar wall. When C-fibers and J- receptors are stimulated by for example inflammation, it may trigger the sensation of breathlessness. Opioids may alter the perception of breathlessness by changing the signals from pulmonary afferent C-fibers. The drive to respiration comes from the brainstem and is influenced by inputs from the cortex, central (brainstem), and peripheral (carotid and aortic bodies) chemoreceptors that register changes in the chemical constituents of blood [37]. Especially rhythm generation is sensitive to opioids and at low doses, opioids cause changes in the respiratory pattern.
Even when opioids can reduce chronic breathlessness, anxiety (due to breathlessness) and improve the quality of life of the patient, still physicians are reluctant to prescribe opioids due to safety concerns [40-42]. These include insufficient knowledge on usage of opioids, lack of treatment guidelines, and fear of respiratory adverse effects. When patients and their caregivers been asked how they experienced the effects of opioids for chronic breathlessness they answered that opioids gave them a sense of calm and relief from severe breathlessness, and improvements in their quality of life [42]. It helped them breath more normally and symptoms of anxiety and depression got better.
Data about respiratory adverse effects of opioids are limited. There is no evidence of significant or clinically relevant for respiratory adverse effects from sustained/extended release opioids in lower doses [43]. Events of respiratory depression is described with high-dose opioids, especially for short acting opioids and parenteral opioids. But for low- dose opioids there is both a clinical and a statistically significant reduction of breathlessness and is therefore a safe and effective treatment when appropriate monitored [43-47]. The risk of respiratory depression seems to be low when treated with opioids for breathlessness, where sustained-release opioids are recommended, because of the doses used. The risk seems to be low even in elderly patients [16]. The risk of respiratory adverse effects is not as a big problem in the palliative care, which will be the case as ILD progresses [48]. There is no association between opioids and increased hospital admission and neither low-dose nor high-dose opioids have association with increased mortality for patients with oxygen-dependent COPD and ILD [31, 49].
Opioids may be under used for chronic breathlessness in advanced respiratory disease.
An observational study from Sweden in patients with advanced, oxygen-dependent COPD found that opioids were not prescribed often for the indication breathlessness, instead, pain was the dominant indication for opioid prescriptions [50]. Despite evidence supporting the use of opioids for breathlessness, they are rarely prescribed for that indication in oxygen-dependent COPD. Due to the questionable effect of benzodiazepines for relieving breathlessness, and the fact that it is still prescribed for this indication, it is
6
interesting to analyze the extent of it´s prescribing in ILD. No study has evaluated the indications for actual prescribed opioids and benzodiazepines for patients in ILD.7
2 Objective
The aim of this study is to evaluate the stated indications and the medication types of dispensed benzodiazepine and opioid prescriptions for patients with severe ILD treated with long-term oxygen therapy in Sweden.
The research questions in this study population are:
• What is the prevalence of treatment with benzodiazepines and opioids?
• Which are the most common stated indications for benzodiazepine treatment and opioid treatment? What proportion of indications are for breathlessness?
• Which are the most common types of benzodiazepines and opioids prescribed?
8
3 Method
3.1 Design
The National Registry for Respiratory Failure (Swedevox) is a register that prospectively includes patients from Sweden starting long-term oxygen therapy (LTOT), with a national coverage of about 85% since 1987. The study population for this observational population-based, longitudinal cohort study were collected from Swedewox based on the following inclusion criteria: patients diagnosed with ILD, aged ≥18 years and starting LTOT between October 1, 2005 and December 31, 2014. If a patient started LTOT more than once during the study period, only the latest LTOT episode was included. Patients who had undergone lung transplantation before starting LTOT were excluded from the study.
This study utilized the following data from Swedewox as a baseline of characteristics:
forced expiratory volume in 1 s (FEV1) % pred, vital capacity (VC) % pred, FEV1/VC, arterial oxygen tension (PaO2) and arterial carbon dioxide tension (PaCO2) while breathing air and oxygen, smoking status, body mass index (BMI), and World Health Organization (WHO) Performance Status. The WHO Performance Status is a scale that grades the patient's overall well-being and activities of daily life. The scale goes from 0-4 where 0 = fully active, able to carry out all pre-disease activities without restriction; 1 = restricted in physically strenuous activity but ambulatory and able to carry out work of a light or sedentary character; 2 = ambulatory and capable of all selfcare but unable to carry out any physically hard work, and up and moving/doing things more than 50% of waking hours; 3 = capable of only limited selfcare, confined to bed or chair more than 50% of waking hours; 4 = completely disabled, incapable of any selfcare, totally confined to bed or chair. A follow up of the baseline characteristics is done after one year and sometimes even after three years [51].
The Swedish Prescribed Drug Register provided information about data on all dispensed prescriptions for the cohort and information about the prescriptions, gender and the age of the study population [52]. The Swedish Causes of Death Register provided the vital status of the study population.
3.2 Types of benzodiazepines and opioids
The categories of benzodiazepines and opioids analyzed in this study are according to Anatomical Therapeutic Chemical (ATC) codes as benzodiazepines (N05BA), weak opioids (N02AA59 and N02AX02), and strong opioids (N02A except weak).
3.3 Method
Random samples (n=2,000 for each treatment) of the participants’ dispensed benzodiazepine and opioid prescriptions during the study period were analyzed in relation to medication type, if it is a regular dosing or as needed, and free-text indication.
A patient was considered to use a drug if it had at least one dispensed prescription. The categorization of indications was conducted by hand, by me, using an excel file. In case of uncertainties, these were discussed with my external supervisor before decisions were made and the work were continued. The indications were examined over time during the observation period but also during the last 6 months of the patients’ lives. All prescriptions from 6 months before LTOT (start of the study) until follow-up (first of death or study end) were included. Prescriptions from 6 months before the patients started LTOT were included to capture drugs that may have been prescribed earlier, but used during the study period, for example as needed. The stated indications categorized as dyspnea and/or pain and/or anxiety and/or other or unknown. Prescriptions could contain more than one indication at once and could have regular dosing and as needed
9
dosing simultaneously. According to Swedish National Board of Health and Welfare (responsible for the management of the Swedish Prescribed Drug Register used in this study) the free-text indication for dose-dispensed prescriptions was not collected until 2014. The defined daily dose (DDD) per prescriptions were analyzed for both benzodiazepines and opioids.3.4 Statistical analyses
The results are presented mainly through descriptive statistics. The baseline of patient characteristics is summarized using mean with standard deviation (SD). For continuous variables who are non-normally distributed, there are median with range or interquartile range (IQR). For categorical variables there are frequencies and percentages. A confidence interval of 95% was used and Chi-square tests were performed with p<0.05 as significant values. The statistical analyses were performed using the software packages Stata, version 14.2 (StataCorp LP; College Station, TX), and SAS, version 9.3 (SAS Institute, Inc., Cary, NC).
3.5 Ethical considerations
The study was approved by the Lund University (Lund, Sweden) research ethics committee (Dnr: 2016/846), the Swedish National Board of Health and Welfare, and the Swedish Data Inspection Board. Individual consent from the patients is not required as the study used de-identified and un-re-identifiable data.
10
4 Results
A total of 1,635 patients with fibrotic ILD started LTOT between October 1, 2005 and December 31, 2014 and were included in the study. Among the study population, benzodiazepines were received by 651 (39.8%) patients and opioids by 710 (43.4%) patients during the study period and/or 182 days before baseline; 373 (22.8%) patients received both benzodiazepines and opioids. The total number of received benzodiazepine and opioid prescriptions were 5,808 and 7,236, respectively.
The random sample of 2,000 opioid and 2,000 benzodiazepine prescriptions was dispensed by 744 patients where benzodiazepine prescriptions included 486 patients, opioid prescriptions 470 patients, and 212 patients had both benzodiazepine and opioid prescriptions.
4.1 Baseline characteristics
Baseline characteristics are presented for all benzodiazepine and opioid prescriptions, as well as for the random sample, to get an overview of the study population's characteristics and to see if anything separates from what is known about ILD. Characteristics were similar between the patients using benzodiazepines and opioids. Most patients had a WHO Performance status 1, for both the whole population and the random sample (Table 1). The biggest comorbidity for both all benzodiazepine and opioid prescriptions were cancer (Table 1), followed by COPD and Diabetes mellitus. Cardiovascular diseases were not present in about 30% of both all benzodiazepine prescriptions and opioid prescriptions (Table 1). Of the benzodiazepine and opioid prescriptions that had cardiovascular disease, it was most common to have three or more diseases. Patients who used benzodiazepines were aged 76±9 years, and patients who used opioids were aged 76±10 years. The patients who used benzodiazepines and opioids were mostly men, except in the random sample where the patients who used opioids were mostly women.
Table 1. Characteristics of patients with oxygen-dependent ILD who were prescribed benzodiazepines and opioids.
Characteristics Whole population Benzodiazepine s
Whole populatio n
Opioids
Random sample Benzodiazepine s
Rando m sample Opioids
n=651 n=710 n=486 n=470
Age (years) 76.2±8.8 75.7±9.5 76.5±8.5 75.7±9.5
Female, n (%) 273 (41.9) 313 (44.1) 214 (44.0) 248
(52.8) FEV1 % pred* 67.1 (21.8) 66.5 (23.3) 67.4 (22.6) 65.5
(24.2)
VC % pred* 56.1 (19.8) 57.1 (21.2) 56.5 (19.5) 56.3
(21.4)
FEV1 / VC* 0.86 (0.56) 0.81 (0.14) 0.86 (.64) 0.81
(0.15) PaO2 Breathing air
kPa* 6.7 (1.03) 6.6 (1.04) 6.7 (0.95) 6.6
(0.94) PaCO2 Breathing
air kPa* 5.1 (0.93) 5.2 (0.94) 5.12 (0.92) 5.2
(0.94) Median (IQR)
follow-up (days) 280 (111-570) 304.5 (118-
646) 329 (135-593) 323 (129- 685) Comorbidities
COPD 140 (21.5) 159 (22.4) 110 (22.6) 117
(24.9)
Cancer 244 (37.5) 286 (40.3) 184 (37.9) 205
11
(43.6) Depression/Anxiety 68 (10.5) 54 (7.6) 65 (13.4) 43 (9.2)
Diabetes mellitus 129 (19.8) 157 (22.1) 94 (19.3) 115 (24.5)
Injury 98 (15.1) 125 (17.6) 74 (15.2) 86 (18.3)
Osteoporosis 36 (5.5) 63 (8.9) 25 (5.1) 48 (10.2)
Renal failure 34 (5.2) 49 (6.9) 27 (5.6) 31 (6.6)
Pulmonary
hypertension 42 (6.5) 63 (8.9) 29 (6.0) 47 (10.0)
GORD 30 (4.6) 42 (5.9) 23 (4.7) 32 (6.8)
Cardiovascular diseases
0 199 (30.6) 197 (27.8) 147 (30.3) 125
(26.6)
1 30 (4.6) 30 (4.2) 21 (4.3) 17 (3.6)
2 173 (26.6) 180 (25.4) 139 (28.6) 116
(24.7)
≥3 249 (38.3) 303 (42.7) 179 (36.8) 212
(45.1) WHO Performance
status
0 25 (3.8) 30 (4.2) 20 (4.1) 21 (4.5)
1 252 (38.71) 291 (41.0) 191 (39.3) 189
(40.2)
2 188 (28.9) 184 (25.9) 136 (28.0) 132
(28.1)
3 116 (17.8) 118 (16.6) 86 (17.7) 72 (15.3)
4 10 (1.5) 9 (1.3) 5 (1.0) 7 (1.5)
Missing 60 (9.2) 78 (11.0) 48 (9.9) 49 (10.4)
Notes: Data presented as mean ± standard deviation, and mean (%), unless stated as something else. *percent of missing in total sample (n=1635): FEV1 % pred (42.1%); VC % pred (42.5%);
FEV1/VC (39.6%); PaO2 breathing air kPa (22.5%); PaCO2 breathing air kPa (22.9%).
Abbreviations: GORD, Gastro-Oesophageal Reflux Disease.
12 4.2 Indications
The indications are presented over time to be able to analyze whether there is a bigger incidence of prescriptions as the disease progresses and gets worse, and at the patients' last months of life. Indications were stated in 52% (n=1041/2,000) of the benzodiazepine prescriptions and in 54.3% (n=1087/2,000) of the opioid prescriptions (Table 2). Of the missing free-text indications, it was mostly for dose-dispensed prescriptions.
Throughout the study period, anxiety was the predominant indication for benzodiazepines and for opioids and it was pain. The indication breathlessness was present in 0.3% of the benzodiazepine prescriptions 6 months before starting LTOT and in 1.0% the last 6 months of LTOT (Table 2). For opioid prescriptions, the indication breathlessness was present in 3.5% 6 months before LTOT and in 4.5% the last 6 months of LTOT.
Table 2. Indications of 2,000 random benzodiazepine and opioid prescriptions among patients with oxygen-dependent ILD.
Indication, n (%)
(CI 95%) Overall
(n=4000) 6 months before
starting LTOT (n=797)
First 6
months of LTOT
(n=1101)
Last 6
months of LTOT
(n=1945) Benzodiazepines (n=2,000) (n=347) (n=586) (n=1036)
Pain 2 (0.1)
(0.0-0.4) 1 (0.3)
(0.0-1.6) 0
(0.0-0.6) 0
(0.0-0.4) Breathlessness 15 (0.8)
(0.4-1.2) 1 (0.3)
(0.0-1.6) 6 (1.0)
(0.4-2.2) 10 (1.0) (0.5-1.8)
Anxiety 1009 (50.5)
(48.3-52.6) 193 (55.6)
(50.4-60.8) 279 (47.6)
(43.6-51.7) 519 (50.1) (47.1-53.1)
Other 53 (2.7)
(1.9-3.4) 12 (3.5)
(1.5-5.4) 18 (3.1)
(1.7-4.5) 35 (3.4) (2.3-4.5) Unknown
indication 959 (48.0)
(45.8-50.1) 147 (42.4)
(37.2-47.6) 297 (50.7)
(46.6-54.7) 501 (48.4) (45.3-51.4)
Opioids (n=2,000) (n=450) (n=515) (n=909)
Pain 1030 (51.5)
(49.3-53.7) 284 (63.1)
(58.7-67.6) 241 (46.8)
(42.5-51.1) 453 (46.5) (46.6-53.1) Breathlessness 70 (3.5)
(2.7-4.3) 4 (0.9)
(0.2-2.3) 27 (5.2)
(3.5-7.5) 41 (4.5) (3.3-6.1)
Anxiety 17 (0.9)
(0.5-1.4) 0
(0.0-0.8) 9 (1.8)
(0.8-3.3) 12 (1.3) (0.6-2.1)
Other 28 (1.4)
(0.9-2.0) 5 (1.1)
(0.4-2.6) 5 (1.1)
(0.3-2.3) 13 (1.4) (0.8-2.4) Unknown
indication 913 (45.7)
(43.5-47.8) 161 (35.8)
(31.1-40.2) 248 (48.2)
(43.8-52.5) 423 (46.5) (43.3-49.8) Notes: 307 of the benzodiazepine prescriptions are in both the first and last 6 months groups. 260 of the opioid prescriptions are in both the first and last 6 months groups. Data presented as number of prescriptions n (%) and lower and higher percentage limit with confidence interval 95%
(CI 95%).
During the last 6 months among the patients who died, the period prevalence of benzodiazepine prescriptions for breathlessness was 1.2% and for opioid prescription it was 4.3% (Table 3). Among the patients who died, anxiety had the highest prevalence for benzodiazepines in both the first 6 months of LTOT and the last 6 months of life, and for opioids pain had the highest prevalence both during the first 6 months of LTOT and the last 6 months of life.
13
Table 3. Stated indications for prescribed benzodiazepines and opioids starting LTOT versus last 6 months of life for patients with oxygen-dependent ILD who died.Indication, n (%) (CI
95%) First 6 months of LTOT
(n=873) Last 6 months of life (n=1584)
Benzodiazepines (n=474) (n=855)
Pain 0
(0.0-0.8) 0
(0.0-0.4)
Breathlessness 4 (0.8)
(0.2-2.1) 10 (1.2)
(0.6-2.1)
Anxiety 215 (45.4)
(40.9-49.8) 413 (48.3)
(45.0-51.7)
Other 13 (2.7)
(1.5-4.6) 27 (3.2)
(2.1-4.6)
Unknown indication 251 (53.0)
(48.5-57.4) 429 (50.2)
(46.8-53.5)
Opioids (n=399) (n=729)
Pain 163 (40.8)
(36.0-45.7) 347 (47.6)
(44.0-51.2)
Breathlessness 16 (4.0)
(2.3-6.4) 31 (4.3)
(2.9-6.0)
Anxiety 6 (1.5)
(0.6-3.2) 11 (1.5)
(0.8-2.7)
Other 5 (1.3)
(0.4-2.9) 7 (1.0)
(0.4-2.0)
Unknown indication 217 (54.4)
(49.5-59.3) 356 (48.8)
(45.2-52.5)
Notes: 307 of the benzodiazepine prescriptions are in both the first and last 6 months groups. 260 of the opioid prescriptions are in both the first and last 6 months groups. Data presented as number of prescriptions n (%) and lower and higher percentage limit with confidence interval 95%
(CI 95%), unless states as something else.
The distribution between WHO Performance status and indications shows significant differences (Table 4). The percentages in Table 4 are for one column for each WHO status. For the benzodiazepine prescriptions, WHO status 1 was most common for all indications. For opioid prescriptions, different WHO statuses were most common for the indications, for example, breathlessness as an indication had WHO status 0 as the most prevalent with 6.8%.
14
Table 4. Indications of benzodiazepine and opioid prescriptions by WHO Performance status in patients with oxygen-dependent ILD.WHO Perfor mance status Indication, n
(%) 0 1 2 3 4 Missing X2
Benzodiazepine s
Pain 0 2
(0.3) 0 0 0 0 <0.001
Breathlessness 1
(1.2) 10
(1.3) 2
(0.4) 2
(0.4) 0 0 <0.001
Anxiety 34 (41.0) 435
(58.1) 256
(46.8) 201
(41.4) 8
(53.3) 75
(62.5) <0.001
Other 0 24
(3.2) 10
(1.8) 9
(1.9) 1
(6.7) 9
(7.5) <0.001
Unknown 48 (57.8) 297
(39.7) 290
(53.0) 276
(56.8) 78
(46.8) 41
(34.2) <0.001 Opioids
Pain 21 (28.4) 395
(50.7) 274
(50.1) 157
(50.2) 15
(53.6) 168
(64.9) <0.001 Breathlessness 5 (6.8) 28
(3.6) 22
(4.0) 10
(3.2) 0 5
(1.9) <0.001
Anxiety 0 8
(1.0) 1
(0.2) 3
(1.0) 0 5
(1.9) <0.001
Other 1
(1.4) 18
(2.3) 6
(1.1) 2
(0.6) 0 1
(0.4) <0.001
Unknown 48 (64.9) 357
(45.8) 264
(48.3) 145
(46.3) 13
(46.4) 86
(33.2) <0.001 Abbreviations: X2, Chi-square test.
15 4.3 Types of benzodiazepines and opioids
The most frequently prescribed benzodiazepines were Oxazepame, and the most frequently prescribed opioids were Oxycodone (Table 5). The benzodiazepine prescriptions were mostly prescribed as an as needed dose, and the opioid prescriptions as a regular dose (Table 5).
Table 5. Benzodiazepine and opioid prescriptions for patients with oxygen-dependent ILD.
Prescriptions Random sample Benzodiazepines
n=2,000
Random sample Opioids n=2,000
X2
Mean DDDs per
prescription (mg) 33 19.7 -
Dosing, n (%) (CI 95%)
Regular dosing 493 (24.7)
(22.8-26.5) 798 (39.9)
(37.8-42.0) <0.001 As needed dosing 888 (44.4)
(42.2-46.6) 648 (32.4)
(30.3-34.5) <0.001 Regular and as
needed 35 (1.8)
(1.2-2.4) 10 (0.5)
(0.2-0.9) <0.001
Unknown 654 (32.7)
(30.6-34.8) 564 (28.2)
(26.2-30.2) 0.002 Benzodiazepine
substances, n (%) (CI 95%)
Oxazepam 1712 (85.6)
(84.1-87.1) - -
Diazepam 139 (7.0)
(5.9-8.2) - -
Alprazolam 123 (6.2)
(5.1-7.3) - -
Other 26 (1.3)
(0.9-1.9) - -
Opioid substances, n (%) (CI 95%)
Codeine - 269 (13.5)
(12.0-14.9) -
Dextropropoxyphene - 97 (4.9)
(4.0-5.9) -
Fentanyl - 132 (6.6)
(5.6-7.8) -
Morphine - 440 (22.0)
(20.2-23.8) -
Oxycodone - 573 (28.7)
(26.7-30.6) -
Tramadol - 384 (19.2)
(17.5-20.9) -
Others - 105 (5.3)
(4.3-6.3) -
Notes: Data presented as number of prescriptions n (%) and lower and higher percentage limit with confidence interval 95% (CI 95%).
Abbreviations: DDDs, Defined Daily Doses; X2, Chi-square test.
16
5 Discussion
5.1 Choice of design
The purpose of this study was to evaluate the indications and medication types of benzodiazepine and opioid prescriptions in severe oxygen-dependent ILD. According to the schematic picture of treatment of breathlessness in Figure 3, LTOT is given in the last step. LTOT indicated in very advanced, end stage COPD, where most patients are affected by severe chronic breathlessness [53]. By choosing patients with LTOT they should very likely have opioids on prescription, and maybe benzodiazepines as well. For this reason, these patients constitute a good target group for the aim of this study. To answer the questions of this study, the data of the cohort’s prescriptions were analyzed together with variables from the Swedish Prescribed Drug Register, which allowed any patterns to emerge.
A registry study is suitable when you want to highlight context and build knowledge. In this study it is interesting to measure whether benzodiazepines are prescribed more often than opioids for breathlessness, as its indication is not as well-founded and has dangerous side effects. Based on the purpose of the study, a registry-based observational study is suitable to investigate the presence and prescriptions of benzodiazepines and opioids in ILD, since registers reflect the reality.
The statistical means for a study should be equivalent to the purpose of the study [54].
There are many factors to consider when to derive scientific evidence, some examples are:
the design of the study, the quality of the measurements, the external evidence for the phenomenon under study, and the validity of assumptions that underlie the data analysis [55]. Therefore, the use of “statistical significance” as “p ≤ 0.05” for justifying scientific claims or conclusions, can lead to twisted results from the scientific process. This is an observational study with observed data and the results can be evaluated for the study population by using interval estimation. A confidence interval indicates the statistical variability and gives a range of interpretable values based on the random error in the measurement, which is not just probability as for P-values; thus, a confidence interval gives more information than a P-value on the uncertainty in a generalization [54, 56-58].
Instead of determining whether a finding is ‘statistically significant’, confidence intervals give values based on the sample data, in which the value for the population may be included (given the absence of residual bias) [56]. Gardner and Altman think that if confidence intervals are appropriate, then they should be used for major findings in a paper. Since the analysis of the indications in this study is done on a random sample, it was more useful to present sample statistics as estimates of how the results would have been if the total number of prescriptions of the study population were studied.
A significance test with P-value could have been done for the indication breathlessness over time to see if it increased, decreased or stayed the same. But in the data, there was an overlap of prescriptions for the first 6 months of LTOT and the last 6 months of LTOT before death, which exceeded 10%. The comparison of the two groups is not two independent quantities, and as a rule of thumb if you have more than 10% of cases in more than one group, comparison is not advised without adjustment in form of a correction factor [59]. Regarding the difficulty in calculating a correction factor, but mostly regarding that it would not have contributed to anything relevant to the study, the significance test for Table 2 and 3 was excluded. When the WHO Performance status 1 was the most common overall for both benzodiazepine and opioid prescriptions, a Chi- square test was performed to analyze it closer since it was not an expected result for those seriously ill. But since this study has a large sample size, P-values will often be <0.05 even when the difference is small. The P-values will therefore show a significant difference between categories, as in Table 4 and Table 5. The random sample contained 48% of benzodiazepine prescriptions and 45.7% of opioid prescriptions lacking their indication, which may influence the values. If these missing indications were found and included in
17
the test, the result would possibly be affected as the indications of e.g. pain and anxiety could increase.Based on this, confidence intervals were used as the main measure for the results in this study, together with descriptive measurements. Since this was an examining study to illustrate and increase knowledge about the use of benzodiazepines and opioids in severe oxygen-dependent ILD, and with no comparisons to any other data due to no related studies in the field, significance tests of differences are not greatly important.
5.2 Main findings
The prevalence of benzodiazepine prescriptions in patients with oxygen-dependent ILD were 39.8% and for opioid prescriptions it was 43.4%. Breathlessness as a stated indication for benzodiazepines and opioids is uncommon in patients with severe oxygen- dependent ILD, also in the last months of life. This finding suggests that many of these patients are undertreated for their chronic breathlessness, both during the progression of the disease but also in the patients last months of life. Prescriptions with the indication breathlessness was overall low with 0.8% (n=15) for benzodiazepine prescriptions and 3.5% (n=70) for opioid prescriptions. Even with the highest value within the 95%
confidence interval, the value was still low for both benzodiazepines (1.2%) and opioids (4.3%) for the indication breathlessness when compared to anxiety (52.6%) for benzodiazepines and pain (53.7%) for opioids, who had the highest values for each treatment. The incidence for prescriptions with breathlessness as an indication were still low in the last 6 months of the patients’ lives with 1.2% for benzodiazepines and 4.3% for opioids. Here, too, the highest value within the 95% confidence interval was low for both benzodiazepines (2.1%) and opioids (6.0%) when compared to anxiety (51.7%) for benzodiazepines and pain (51.2%) for opioids. The results of this study that neither benzodiazepines nor opioids are prescribed often for the indication breathlessness in ILD (overall 3.5% and 0.8% respectively), which indicates undertreatment of patients with severe ILD in the end of life, are valid because of the national and registry-based design and the evidence of opioids in the relief of breathlessness [43-47]. This result complies with other studies that breathlessness in ILD is undertreated at the end of life [21, 22, 27, 31]. Palliative and end-of-life care is of crucial importance for these severely ill patients with a high symptom burden and not enough specialized palliative care may lead to worse control and recognition of suffering from symptoms, which the result of Ahmadi et al.
supported where lower use of structured symptom assessments was seen for patients with ILD in the end-of-life [27]. Since opioids should be given before LTOT (Figure 3), the expectation was that the prevalence of breathlessness as an indication would been higher.
Compared to the other study from Sweden who analyzed the safety of benzodiazepines and opioids in ILD, the proportions of prescriptions differed 27.8% for benzodiazepines and 27.4% for opioids, which is a significant difference [31]. Since that study did not examine indications, only the total amount of prescriptions can be compared. The reason behind this difference may be that the inclusion criteria for age was different and may also be of chance due to the random sampling. The result of this study also differs from another study in the UK where the palliative care of ILD were analyzed in the last year of life [21]. Their result was much lower for patients receiving benzodiazepines (18%), but similar regarding opioids (49%). Since this is comparing two different countries there may be differences in the regulatory framework for prescription of drugs. For example, both benzodiazepines and opioids are recommended for breathlessness in idiopathic pulmonary fibrosis (IPF) in the UK [60].
Of the study population who died, anxiety was the most common indication for benzodiazepines and pain was the most common indication for opioids. Breathlessness as an indication were uncommon during the last 6 months of the patients’ lives, it was present in 1.2% of the benzodiazepine prescriptions and only in 4.3% of the opioid prescriptions. With the highest value in the 95% confidence interval, the indication breathlessness was still low for both benzodiazepines (2.1%) and opioids (6.0%) regarding
18
that at the end of life, symptom relief is of great importance in reducing the patient's suffering. Since opioids are part of the treatment regimen in the case of breathlessness in palliative care, the low results of opioid prescriptions indicate undertreatment in the last months of life [55]. As shown in the study where lung cancer and ILD were compared in Sweden, where patients with ILD had lower rates of complete relief from breathlessness and were less frequently prescribed as needed medication, which is often directed at all symptoms [26]. Ahmadi et.al. suggests that patients dying with oxygen-dependent ILD receive lower quality end-of-life care and experience more breathlessness when compared to patients dying of lung cancer.One possible explanation for the fact that prescribing of opioids were higher during the study period than prescribing of benzodiazepines, may be that opioids actually have clinical evidence of improving breathlessness while benzodiazepines do not. When physicians are faced with choosing treatment for breathlessness, they most probably rely on the clinical evidence available and therefore choose opioids rather than benzodiazepines.The number of patients treated with opioids were still low whichmay be because that the physicians ignore breathlessness as a therapeutic target, but if they note it as one, the uncertainty of treatment prevents them from prescribing opioids. This finding, that opioids are prescribed more often for breathlessness than benzodiazepines, contradicts the fears that the prescription of benzodiazepines was greater than opioids despite their lack of efficacy and side effects. The fact remains that benzodiazepines are still prescribed for this indication in some patients despite lack of evidence. This could be the case since they may be prescribed after opioids are prescribed with no effect, and then tested as a secondhand choice. Another explanation is that benzodiazepines are prescribed because of the anxiety associated with breathlessness, and thus thinking they will relieve the anxiety and breathlessness without knowing what the evidence base says.
Physicians who find that benzodiazepines are effective may continue to prescribe them to other patients, which is another possible explanation for their prescription for breathlessness.
The most predominant indication throughout the study period was anxiety for benzodiazepine prescriptions and pain for opioid prescriptions, which was expected since anxiety is one of the main indications for benzodiazepines and pain is for opioids [29, 36].
There may be a greater number of breathlessness because of any double indication but where only one is written on the prescription. For example, since anxiety and breathlessness are associated with each other [13, 48, 61-65], there may be a possibility that the patient has both conditions but that the physician has written only one of them as an indication on the prescription.
All dosing regimens had significant differences for both benzodiazepine and opioid prescriptions. Benzodiazepines were mostly given as needed, which may be because of their main indication, anxiety [32]. Anxiety may come every now and then but can also last for a long time [66]. The designated dosage is designed according to the patient's needs and in case of anxiety it may be needed both as regular dosage and if needed based on the nature of the anxiety. Opioids are usually given for pain and this may be linked to the result that it is given most frequently as regular dosing [36, 50]. If the patient suffers from constant pain, a regular dosing is needed to achieve pain relief. Benzodiazepines have a greater incidence in prescriptions who have both a regularly dosing and as needed (1.8% (n=35)) than opioids (0.5% (n=10)). Each dosage is individual and designed according to the patient's needs. Therefore, the reason behind this difference cannot be determined. Anxiety can, however, express itself both vaguely and strongly as pain can have breakthrough episodes.
The most frequently prescribed benzodiazepine in the random sample was Oxazepame (85.6%), and the most frequently prescribed opioid was Oxycodone (28.7%), Morphine (22%), Tramadol (19.2%), Codeine (13.5%), Fentanyl (6.6%) and Dextropropoxyphene (4.9%). The results clearly show that opioids have a greater spread of several substances while benzodiazepines are prescribed in about 85% of cases as only one substance. One
19
possible explanation for the high incidence of Oxazepame may be due to that is the only benzodiazepine that is recommended for the elderly [67]. It is also first-hand treatment for short-term anxiety. The high presence of Oxycodone (morphine derivative) and Morphine may be because Morphine is the recommended opioid for breathlessness in palliative care [61]. It is difficult to name the reason for the high incidence of Tramadol when it should not be prescribed to the elderly due to side effects like dizziness [36]. But it can be prescribed to elders in special cases and the side effects can be controlled if it is given as a slow-release preparation and with slow dose titration [38]. These results of the most common substances comply with Ahmadi et al. for the order of opioids [50]. The only difference is that they had both Oxycodone and Tramadol as the most common substances. Their study only analyzed opioid prescriptions for patients with oxygen- dependent COPD.Baseline characteristics were similar between participants who were prescribed benzodiazepines and opioids, where there was an overlap of 212 patients using both benzodiazepines and opioids. The most common WHO Performance status for all benzodiazepine and opioid prescriptions overall (Table 1) were status 1, which means that the patient is restricted in physically strenuous activity but ambulatory and can perform work of light or sedentary character. The most common Who Performance status for opioid prescriptions with the indication breathlessness were 0, which means that the patient is fully active, and able to carry out all pre-disease activities without restriction.
These results of low scores on the WHO Performance statuses were not expected for those seriously ill. Since the inclusion criteria for this study were long-term treatment with oxygen, and since oxygen treatment are given for chronic lung failure, the expectation were high scores for this study population. A possible explanation for this unexpected result may be that the clinician overestimated the patient's function when grading,
or just copied over from the previous consultation without a reassessment.
But may also be the result of patients with a better physical status moving more and becoming more breathless from the exertion and therefore being treated more.The results showed that the patients were mostly men for all prescribed benzodiazepines and opioids. This was expected due to a greater incidence of ILD in men [8]. But in the random sample of the opioid prescriptions, there were mostly female (52.8%). One possible explanation for this is the result of chance from random sampling since the overall opioid prescriptions consisted mostly of men. Another possible explanation is that this is a sample of all the prescriptions and if patients have more than one prescription, they become over-represented. So, women may be given more prescriptions than the results show. This result is different compared to Bajwah et al. where the majority were women for patients receiving both benzodiazepines and opioids compared with patients without any of these prescriptions [31]. Their study population was also collected from Swedevox during 2005-2014, but one of the inclusion criteria was patients aged over 45 years and this study has patients aged over 18 years. This may explain the different result, but it can be a result from the chance by the random sampling.
5.3 What this study adds
The findings in this study forms the basis for future research as a baseline to evaluate prescribing of both benzodiazepines and opioids to patients with ILD. It will also contribute to increased knowledge for this study population as no information was available for prescribing benzodiazepines and opioids and their indications for patients with ILD. The results of this study suggest undertreatment of patients with ILD, which creates a clearer picture of the patients' reality. The study did not analyze the degree and nature of breathlessness but the finding that ILD is undertreated indicates that these patients are at greater risk of suffering. It may also contribute to identify possible improvement for clinicians regarding treatment of ILD, improving the overall well-being of patients. An example is the finding that many prescriptions lack indication, not just for dose prescriptions, which gives a chance for improvement. Since Tramadol is not
20
recommended for the elderly, this study may be used as the baseline for the development of Tramadol prescription to the elderly.5.4 Strengths and limitations
One of the strengths of this study was its national population-based design with complete longitudinal data on all dispensed benzodiazepine and opioid prescriptions. The use of a national registry (Swedevox) with nation-wide coverage likely yields a high external validity to patients in Sweden and similar settings, and no patient was lost to follow up.
This unique national data in Sweden makes it possible to follow these patients with advanced ILD over time, which is otherwise difficult to follow due to the lack of information collected and stored and, among other things, measured values. Another strength was that this study had a large random sample of dispensed prescriptions which makes the result likely representative of all dispensed prescriptions for the cohort. The random sample evaluated presence of relevant and concurrent indications regarding ILD, which enables comparisons across studies.
For the second and third objectives, the unit of analysis was prescriptions. For the random sample of prescriptions, one patient’s prescriptions might have been chosen several times, while another patient’s prescriptions were only chosen once or not at all.
This may introduce a skew in the analysis of patient data and the results, which would have been prevented by including all patients' benzodiazepine and opioid prescriptions, but that had been too many prescriptions to analyze by hand. Hence, a randomized sample of 2,000 prescribed benzodiazepine and opioid prescriptions were made, respectively. Another limitation was that almost half of the free-text indications for prescriptions were missing. Since this is a register study, the result reflects the clinical reality. The finding that indications of prescriptions are missing (not only for dose- dispensed prescriptions) may contribute to improvement in clinical practice. A prescription should contain information about indication and aim of the treatment, especially for sick elderly people who often get too many and unnecessary prescriptions printed and when informal caretakers are involved. The indication contributes to the safety of the therapy and may have a positive effect on compliance for the patients. If the indications for dose-dispensed prescriptions had existed, the indication of breathlessness itself would not have increased drastically as the overall figures were very low for both benzodiazepines and opioids, even in the last months of the patients' lives. For this reason, and along with other studies whose results indicate that breathlessness is undertreated, the prevalence of breathlessness as an indication is unlikely to increase significantly [21, 22, 27, 31]. To include the dose-dispensed prescriptions made it possible to measure the prevalence of benzodiazepines and opioids in patients with severe oxygen- dependent ILD. The P-values from the Chi-square test would however be affected if the prescriptions with the missing indications were found and included. It would be interesting to investigate the same questions, though with data collected after 2014, for which the indication for these prescriptions should be included.
5.5 Implications
It would be desirable that the high prevalence of prescribed Tramadol be reduced for this seriously ill and elderly population as it is not recommended due to side effects that may pose a danger. However, it may have declined since the end of the study period in 2014.
To be able to introduce evidence-based opioid therapy for breathlessness in advanced ILD, more research is needed that provides additional evidence and information that enhances opioid treatment reliability.
21
6 Conclusion
In severe oxygen-dependent ILD, benzodiazepines and opioids are prescribed in about 40 percent of the cohort, but they are rarely prescribed for breathlessness, even in the last months of life. Which indicates that these patients are undertreated for their breathlessness and are at greater risk of suffering. The most common stated indication was anxiety for benzodiazepines and pain for opioids. The most common prescribed substances were Oxazepame and Oxycodone.
22
7 Acknowledgement
First of all, I would like to extend a big thank you to my external supervisor Magnus Ekström who guided me on how to write this work. Thank you for the quick and solid feedback and all the rewarding ideas. Without you, this work would not been possible. I would like to thank my internal supervisor Maria Sjölander as well, who provided great feedback and made sure that the work met the objectives of this course. And last but not least, a thank you to my partner who supported me during this work.
23
8 References
1. Bourke SJ. Interstitial lung disease: progress and problems. Postgrad Med J. 2006 Aug; 82(970): 494–499.
2. WebMD. Interstitial Lung Disease (ILD) [Internet]. WebMD, LLC; [reviewed 2019-10-08]. Retrieved from: https://www.webmd.com/lung/interstitial-lung- disease#2
3. Wang H, Naghavi M, Allen C, Barber MR, Bhutta AZ, Carter A, et al. Global, regional, and national life expectancy, all-cause mortality, and cause-specific mortality for 249 causes of death, 1980-2015: a systematic analysis for the Global Burden of Disease Study 2015. Lancet. 2016 Oct 8;388(10053):1459-1544.
4. https://www.ncbi.nlm.nih.gov/pubmed/?term=Which+patients+with+advanced +respiratory+disease+die+in+hospital%3F+A+14-year+population-
based+study+of+trends+and+associated+factors
5. Mayo Clinic. Interstitial lung disease [Internet]. Mayo Foundation for Medical Education and Research (MFMER); 2017 [published 2017-07-21]. Retrieved from:
https://www.mayoclinic.org/diseases-conditions/interstitial-lung- disease/symptoms-causes/syc-20353108
6. Choi W, Dauti S, Kim JH, Park HS, Park SJ, Lee WC. Risk factors for interstitial lung disease: a 9-year Nationwide population-based study. BMC Pulm Med. 2018;
18: 96
7. American Lung Association. Interstitial Lung Disease (ILD) [Internet]. Chicago.
Retrieved from: https://www.lung.org/lung-health-and-diseases/lung-disease- lookup/interstitial-lung-disease/
8. Speakman L, Walthall H. Assessment and management of refractory breathlessness in interstitial lung disease. Br J Community Nurs. 2017 Sep 2;22(9):434-439.
9. Kalchiem-Dekel, Jeffrey R. Galvin, Allen P. Burke, Sergei P. Atamas, Nevins W.
Todd. Interstitial Lung Disease and Pulmonary Fibrosis: A Practical Approach for General Medicine Physicians with Focus on the Medical History. J Clin Med.
2018 Dec; 7(12): 476.
10. Läkartidningen. Nytt hopp för patienter med idiopatisk lungfibros [Internet].
Stockholm: Läkartidningen; 2013. Retrieved from:
http://www.lakartidningen.se/Functions/OldArticleView.aspx?articleId=19278 11. Riskförbundet HjärtLung. Vårdprogram för Idiopatisk lungfibros [Internet].
Stockholm; 2012. Retrieved from: https://www.hjart- lung.se/globalassets/riksforbundet-dokument/diagnostik/lungor/nationellt- vardprogram-idiopatisk-lungfibros.pdf
12. Janusinfo. Interstitiella lungsjukdomar [Internet]. Stockholm: Janusinfo;
[updated 2019-10-02]. Retrieved from:
https://janusinfo.se/behandling/akutinternmedicin/lungsjukdomar/lungsjukdo mar/interstitiellalungsjukdomar.5.304d30c1612954524571a8.html
13. Ekström PM, Abernathy PA, Currow CD. The management of chronic breathlessness in patients with advanced and terminal illness. BMJ. 2015 Jan 2;350.
24
14. Brigham Health. Interstitial Lung Disease (ILD) [Internet]. Boston. Retrieved from: https://www.brighamandwomens.org/lung-center/diseases-and- conditions/interstitial-lung-disease-ild15. Manning HL, Schwartzstein MR. Pathophysiology of Dyspnea. N Engl J Med. 1995 Dec 7; 333:1547-1553.
16. Parshall MB, Schwartzstein RM, Adams L, Banzett RB, Manning HL, Bourbeau J, et al. An official American Thoracic Society statement: update on the mechanisms, assessment, and management of dyspnea. Am J Respir Crit Care Med. 2012 Feb 15;185(4):435-52.
17. Currow DC, Plummer JL, Crocket A, Abernathy PA. A Community Population Survey of Prevalence and Severity of Dyspnea in Adults. J Pain Symptom Manage.
2009 Oct;38(4):533-45.
18. Bolton EC, Bevan-Smith FE, Blakey DJ, Crowe P, Elkin LS, Garrod R, et al. British Thoracic Society guideline on pulmonary rehabilitation in adults: accredited by NICE. BMJ. 2013;68:ii1–ii30.
19. Abernethy AP, McDonald CF, Frith PA, Clark K, Herndon JE 2nd, Marcello J, et al. Effect of palliative oxygen versus room air in relief of breathlessness in patients with refractory dyspnoea: a double-blind, randomised controlled trial. Lancet.
2010 Sep 4;376(9743):784-93.
20. Ekström M. Dyspné, kronisk [Internet]. Internetmedicin; [updated 2019-04-01].
Retrieved from: https://www.internetmedicin.se/page.aspx?id=5944
21. Bajwah S, Higginson IJ, Ross JR, Wells AU, Birring SS, Patel A, et al. Specialist palliative care is more than drugs: a retrospective study of ILD patients. Lung.
2012 Apr; 190: 215–220.
22. Bajwah S, Higginson IJ, Ross JR, Wells AU, Birring SS, Riley J, et al. The palliative care needs for fibrotic interstitial lung disease: a qualitative study of patients, informal caregivers and health professionals. Palliat Med. 2013 Oct;27(9):869-76.
23. Johnson MJ, Yorke J, Hansen-Flaschen J, Lansing R, Ekström M, Similowski T, et al. Towards an expert consensus to delineate a clinical syndrome of chronic breathlessness. Eur Respir J. 2017 May 25;49(5).
24. Carvajalino S, Reigada C, Johnson MJ, Dzingina M, Bajwah S. Symptom prevalence of patients with fibrotic interstitial lung disease: a systematic literature review. BMC Pulm Med. 2018 May 22;18(1):78.
25. Swigris JJ, Kuschner WG, Jacobs SS, Wilson SR, Gould MK. Health-related quality of life in patients with idiopathic pulmonary fibrosis: a systematic review.
Thorax. 2005 Jul;60(7):588-94.
26. Matsunuma R, Takato H, Takeda Y, Watanabe S, Waseda Y, Murakami S, et al.
Patients with End-stage Interstitial Lung Disease may have More Problems with Dyspnea than End-stage Lung Cancer Patients. Indian J Palliat Care. 2016 Jul- Sep; 22(3): 282–287.
27. Ahmadi Z, Wysham GN, Lundström S, Janson C, Currow CD, Ekström M. End-of- life care in oxygen-dependent ILD compared with lung cancer: a national population-based study. Thorax. 2016 Jun;71(6):510-6.
