Appropriateness of care

Appropriateness of care – how appropriate care is – may be defined as ’the degree to which provided healthcare is relevant to the clinical needs, given the current best evidence’.

Appropriateness of care can be assessed through different methods, the most robust one of which consists in examining how well medical practice complies with existing clinical recommendations. Another commonly used method is analysing the geographical variability of care practices.

In this report, six indicators of appropriateness affecting acute and chronic care have been selected; they relate to the application of clinical recommendations on diabetes patient monitoring (QA-1, QA-2), antibiotic or antidepressant prescription habits (QA-3 to QA-5) (antidepressants are developed in the Mental Healthcare section), use of inappropriate imaging techniques (QA-6) and inappropriate screening of breast cancer (QA-7).

Summary of the indicators on appropriateness of care​​

(ID) indicator		Score	BEL	Year	Fla	Wal	Bru	Source	EU-15 mean
Primary care – patients with chronic disease (guidelines)
QA-1	Proportion of adult diabetics with appropriate follow-up (% of diabetic patients under insulin) (1)		30.2	2016	32.5	26.0	31.1	IMA-AIM	-
QA-2	Proportion of adult diabetics with appropriate follow-up (% of diabetic patients not under insulin, aged 50+) (1)		11.0	2016	11.3	10.2	13.1	IMA-AIM	-
Primary care – prescribing patterns (guidelines)
QA-3	Use of antibiotics (total DDD/1000 pop/day)		27.7	2016	26.4	30.3	23.4	Farmanet	20.1 (2)
QA-4	Use of antibiotics at least once in the year (% of population)		39.6	2016	38.5	43.7	35.3	IMA-AIM	-
QA-5	Use of antibiotics of second intention (3) (% total DDD antibiotics)		51.8	2016	49.3	56.8	47.9	IMA-AIM	-
Inappropriate medical imaging
QA-6	Spine imaging (X-ray, CT scan, MRI per 100 000 population)		10620	2017	9944	12314	9436	RIZIV-INAMI	-
Cancer overscreening
QA-7	Breast cancer screening outside age target group (% women aged 41-49)		35.4	2016	26.3	49.0	46.7	IMA-AIM	-

(1) Appropriate follow-up is defined as patients receiving regular retinal exams and blood tests (glycohemoglobin, glycaemia, lipid profile and microalbuminuria), (2) OCDE Health Statistics 2018, (3) Antibiotics of second intention are: amoxycilline with clavulanic acid, macrolides, cephalosporins and quinolones.

Some indicators analysed in other sections of this report may also be interpreted in terms of (in)appropriateness of care when their results are excessive compared to recommendations:

• Among mental healthcare indicators: rate of involuntary committals in psychiatric hospital wards (MH-4) or visits to emergency rooms for social, mental or psychic reason(MH-5), and proportion of patients with short duration (< 3 months) of antidepressants treatment (MH-8).
• Among mother and newborn care indicators: number of antenatal consultations for low-risk pregnancies (MN-10), repeated toxoplasmosis C screening during pregnancy (MN-8), induction rate (MN-4), proportion of vaginal births following a previous C-section (MN-6), episiotomy rate (MN-5), very preterm births in hospital without NICU (neonatal intensive care unit, MN-7), Number of newborn screening tests performed outside the required deadlines (‘heel-stick’ test) (a sub-indicator of MN-9).
• Among care for the elderly indicators: prescription of anticholinergic drugs to elderly people (ELD-10) or prescription of neuroleptics in and out of residential facilities for elderly people (ELD 11 and 12).
• Among end-of-life care indicators: proportion of patients who died within one week after start of palliative care (EOL-2), proportion of patients who received chemotherapy in the last 14 days of life (EOL-3).

Diabetes patient monitoring (QA-1 and QA-2)

Diabetes is a chronic disease which is characterised by excessively high glucose levels in the blood. If incorrectly treated, people with diabetes are at a high risk of developing cardiovascular diseases (myocardial infarction, stroke) or kidney failure. Uncontrolled diabetes also increases the risk of visual impairment or lesions to the nerves and blood vessels which can lead to chronic wounds on the feet, or even to amputation. In Belgium, it is estimated that diabetes affects approximately 6% of the population (types 1 and 2 combined – 2017 figure). 
Monitoring a person with diabetes consists in checking their blood glucose levels at least every 3 months, their blood glycosylated haemoglobin (HbA1c) levels (measuring the glycaemic ‘load’ in the blood) at least twice per year, as well as their microalbuminuria (a measurement of kidney function) and a blood lipid profile test at least once a year. An ophthalmologist also needs to perform a funduscopy (fundus of the eye examination) once a year.
This indicator measures the number of individuals with diabetes who have had each of these five tests done at the recommended frequencies during the last 15 months. It is calculated separately for diabetics taking insulin (QA-1) and for diabetics (over 50 years of age) who are treated with anti-diabetic medicines other than insulin (QA-2).

RESULTS

• Among the insulin-treated diabetes patients, 30.2% had done all of the 5 tests selected for assessing the quality of diabetes monitoring during the last 15 months.
• The glycosylated haemoglobin and cholesterol measurements were well covered, but the blood glucose and microalbuminuria tests were done less frequently. The annual rhythm of visits with an ophthalmologist was less well observed.
• In diabetes patients (over 50 years of age) treated with anti-diabetics other than insulin, the 5-test monitoring was only done for 11%.
• Here too, the least well observed tests were blood glucose and microalbuminuria as well as the annual visit with an ophthalmologist.
• Patients aged 75 years and over received better monitoring at home than in residential facilities.
• There were few regional differences for both sub-groups.

Link to technical datasheet and detailed results

Prescription of antibiotics (QA-3, QA-4, QA-5)

It has become a well-known fact that antibiotic consumption facilitates the development of resistant bacteria in humans and animals or in the environment, and that the current upsurge in such resistant bacteria today is a serious threat to all. For several years, the clinical recommendations have been unanimous: these medicines should only be prescribed when there is a verified need for them and when their effectiveness is scientifically proven. This, however, is far from being the case. For example, in the event of a viral infection (influenza, pharyngitis, etc.), antibiotics have no effect; yet they are massively prescribed in Belgium.

In addition, some powerful antibiotics such as quinolones and cephalosporins should be reserved for ‘second-line’ treatments (i.e. if a first antibiotic treatment has not been effective), so that the chances of curing serious and/or resistant infections are not ‘wasted’.

In May  2015, the World Health Organisation (WHO) approved a worldwide action plan for combating antibiotic resistance. In Europe, a European Surveillance of Antimicrobial Consumption Network (ESAC-Net) monitors the consumption of these medicines on an ongoing basis. Since 1999, Belgium has established a committee for antibiotic policy coordination (Belgian Antibiotic Policy Coordination Committee -, BAPCOC) which has developed a new strategy for 2015-2019 and defined several goals. One of these goals is to cut the number of antibiotic prescriptions by half by 2025.

In this report, we have selected 4 indicators related to antibiotic prescriptions on an outpatient basis:

• Total volume of antibiotics prescribed, measured in Defined Daily Doses (DDD) per 1,000 population per day (QA-3)
• Proportion of patients having received at least one antibiotic prescription during a given year (QA-4)
• Proportion of second-line antibiotic prescriptions (quinolones, cephalosporins, amoxicillin-clavulanate and macrolides) compared to the total prescribed volume (QA-5)

RESULTS
Total volume of antibiotics prescribed (QA-3)

• The total volume of antibiotics prescribed on an outpatient basis is 27.7 DDD per 1000 population per day (2016). This is much more than in some other European countries, such as the Netherlands, where the number is only 9.7 DDD.
• Wallonia has the highest consumption level, with over 30 DDD per 1000 population per day. Brussels has the lowest consumption level: 23.4 DDD. Flanders is in between the two.

Figure 4 - DDD antibiotics delivered in the ambulatory sector, DDDs per day per 1000 insured population, standardised relative variations around the national mean, per district, 2016
Data source: Pharmanet (INAMI-RIZIV)

Figure 5 - Overall volume of antibiotics, DDD AB per day per 1000 inhabitants, international comparison (2016 and 2000-2017)
Data source: OECD health statistics 2018

Proportion of patients having received at least one antibiotic prescription during a given year (QA-4)

• In 2016, 40% of the population received at least one antibiotic prescription (43.7% in Wallonia, 38.5% in Flanders and 35.3% in Brussels).
• Among the elderly people (aged 75+ years) living in residential facilities, 62% received at least one antibiotic prescription in 2016, versus 44.4% of individuals of the same age not living in a residential facility.
• Among all provinces, Hainaut had the highest number of people exposed to antibiotics (45.2%).

Proportion of second line antibiotic prescriptions (QA-5)

• Even though it has slightly improved over time, the proportion of second-line antibiotics remains fairly high in Belgium: 52% in 2016 (60% in 2006). This indicator even showed poor results in children (35%) and throughout the country. In the Netherlands, the same indicator was 16% for the same year. 

Link to technical datasheet and detailed results

Inappropriate use of medical imaging exams (QA-6)

Exposure to ionising radiation (such as X-rays, for example) can cause cancer. However, in Belgium, this exposure is particularly high in the medical sector, essentially due to the excessive use of medical imaging exams (X-rays and CT scans). These exams are very often unnecessary and represent alone a very significant share of received radiation doses. As an example, a simple lumbar spine X-ray is equivalent to 680 days (nearly 2 years) of natural radiation, and a lumbar spine CT scan represents 1825 days (5 years) of natural radiation!  National campaigns organised by the FPS [Belgian Federal Public Service] and INAMI – RIZIV have been able to reverse this trend between 2009 and 2012 and to bring figures back to their 2006 level, but the radiation doses received by patients remain in the top tier of European countries.

Note: Magnetic resonance imagery (MRI) does not use X-rays or any other type of hazardous radiation, but its cost and the lack of MRI device availability are a problem.

In this report, we have selected as a primary indicator the number of prescriptions for spine examinations (X-rays, CT scans and MRIs), with a particular focus on lower back examinations, to the extent of the data available. Indeed, with reference to the national (KCE) and international (NICE/UK National Institute for Health and Care Excellence) clinical recommendations, as well as recommendations from the Belgian Radiology Society, these exams are most of the time useless, since their results generally have no influence on the suggested treatment. The use of medical imaging for the spine is therefore a particularly relevant indicator of the (in)appropriateness of care in Belgium.

As an indication, we have also counted the number of CT scanners and MRI devices available in Belgian hospitals.

RESULTS 

• The use of medical imaging (all types of examinations combined) for the spine has overall been decreasing by 2% per year since 2007;
• The number of spine X-rays has been decreasing by nearly 8% per year since 2015;
• The number of CT scans has stopped increasing since 2015. However, some districts have recorded nearly three times more examinations than others (average ratio: 1.8).
• The use of magnetic resonance imaging (MRI) has continued to increase, but at a slower rate: by 6.8% per year in 2007 and by 3.8% per year in 2017.
• The total average radiation rate of the population has decreased by 1% between 2015 and 2017.

Number of devices

• The number of MRI devices has remained virtually unchanged since 2014, which represents a slightly lower density rate than the EU-15 average (12.6 MRI units per million population in Europe versus 11.7 in Belgium in 2016). Eighteen new units will be added in 2021 and 2022 in order to limit the use of CT scans.
• The number of CT scanners has also been relatively stable.

Figure 6 - Spine medical imaging: overall consumption (2007-2017)
Data source: N documents (INAMI-RIZIV)

Figure 7 - Spine medical imaging: relative variations around the national mean per district (2015-2017, standardised)
Data source: N documents (INAMI-RIZIV)

Figure 8 - Spine medical imaging: x-ray consumption (2007-2017)
Data source: N documents (INAMI-RIZIV)

Figure 9 - Spine medical imaging: MRI consumption (2007-2017)
Data source: N documents (INAMI-RIZIV)

Figure 10 - Spine medical imaging: CT scans consumption (2007-2017)
Data source: N documents (INAMI-RIZIV)

Figure 12 - Rate of imaging spine use (per 100 000 insured population) by region and by province (2017)
Data source: INAMI-RIZIV

Figure 13 - Number of MRI units in hospitals per million inhabitants: international comparison (2000-2016)
Data source: OECD health statistics

Figure 14 - Number of MRI units in hospitals per million inhabitants: international comparison (2016)
Data source: OECD health statistics

Link to technical datasheet and detailed results

Inappropriate breast cancer screening tests (QA-7)

This indicator is developed in the section of preventive care indicators.