Although Pirkanmaa and MIKSTRA cannot be compared directly it seems obvious that the ten per cent lower proportion of patients who get antimicrobial prescription for an infection in MIKSTRA compared to Pirkanmaa reflects a true change towards more restrictive prescribing. Also, according to the Finnish Statistics on Medicines, the annual antibiotic sales declined 13 per cent from 1994 to 1998 i.e. before the MIKSTRA programme started and there has been an additional decrease of 5.9% since then (Figure 3, page 29).
It is difficult to assess whether MIKSTRA has contributed to this decline or not. In the same period of time there have been even bigger declines reported in other countries in antibiotic prescriptions for acute respiratory tract infections in children (Ashworth et al., 2004, Finkelstein et al., 2003, Kozyrskyj et al., 2004, McCaig et al., 2002, Unsworth & Walley, 2001). These studies also found a remarkable decrease in the number of consultations for mild respiratory tract infections. Something similar may also be going on in Finland, thus explaining why the declining trend in national statistics started already before MIKSTRA. It may also partly explain the decrease in the number of patients in the MIKSTRA study. The authors of the articles from UK, USA and Canada speculate as to whether the decline in the frequency of consultations is due to patients' increased awareness of the nature of illness of mild respiratory tract infection and a subsequent higher threshold before consulting the doctor in connection with it.
The Pirkanmaa study brought the issue of unnecessary prescribing and antibiotic resistance into public debate in Finland and MIKSTRA has helped to keep it there since then, which may have had an influence on patients' attitudes.
Prescribing of antimicrobials seemed to be, in general, more restrictive in small, rural MIKSTRA health centres than in urban area, unlike in an Australian study from 1990s where the opposite appeared to true (McManus et al., 1997).
The same phenomenon is evident also in the Finnish archipelago - the Åland Islands - where medicines consumption is characterised by low pharmaceutical
92
consumption and costs in most drug groups (including antibiotics), conservative pharmaceutical choices and characteristic patterns (Lahnajärvi et al., 1997).
There might be several possible explanations for this. The incidence of infections may be lower in thinly populated areas. More people are working at home in rural, agricultural areas. They don't need to go to see a doctor for minor infections to get sick leave or it is laborious do so. Families need to take their small children less often to day-care-centres, where children would get more infections, which they would again transmit to their family members (Forssell et al., 2001, Möttönen & Uhari, 1992, Nurmi et al., 1991, Pönkä et al., 1991). On the other hand, distances in Finland, unlike perhaps in Australian rural areas, are perhaps not as long as it would push the patient to demand and the doctor to prescribe antibiotics 'just-in-case'. Furthermore, when doctors and patients are familiar with each other it is easier for the doctor to assess the need for antibiotic of his/her patients and for the patient to rely on the doctors' decision (personal observation). This again helps to avoid 'just-in-case' -treatments and makes it easier to favour even the watchful waiting strategy. And finally, in a rural group practice with stable personnel, doctors' performance may show more coherence and also, due to the distances, will experience less exposure to face-to-face pharmaceutical marketing (Lahnajärvi et al., 1997).
In our study amoxicillin was by far the most commonly used antibacterial followed by macrolides and cephalosporins, while the use of amoxicillin and cephalosporins was rather even in the national data. The differences in the ranking-order do not contradict each other. National sale statistics cover the whole year while our data-collection was a point prevalence survey situated in the high season of respiratory tract infections. Sales of cephalosporins, which are mainly used for skin infections, have hardly any seasonal variation (Figure 18) and they are often used for longer periods of time than antibiotics used for respiratory tract infections. Thus cephalosporins accumulate more in national statistics, while antibiotics such as amoxicillin, which are used mainly for respiratory tract infections, are more represented in our data.
Increasing use of amoxicillin, which was found in our data, was a desired trend and in line with the recommendations, as it is the recommended drug of choice in most respiratory tract infections. Increasing use of amoxicillin with clavulanic acid, on the other hand, does not have a sound basis in the development of the resistance situation and this clearly necessitates some kind of intervention. Unnecessarily prescribing more expensive antibiotics with more side-effects does not benefit the patient and does not constitute good quality of care.
The continuously decreasing proportion of penicillin-V is a somewhat undesirable and unnecessary development. As the most common causative agents in community acquired respiratory tract infections are on the other hand viruses and on the other the potentially dangerous bacteria, Streptococcus pneumoniae and pyogenes, antibacterial treatment should, in most cases, cover the latter two. Penicillin has maintained its effect in these pathogens in Finland and could obviously be used more often.
93 Figure 18. Number of reimbursed prescriptions of antimicrobials by month from
January 1998 to March 2003 (Source: National prescription register)
0 10 000 20 000 30 000 40 000 50 000 60 000 70 000
I/98 II/98 III/98 IV/98 I/99 II/99 III/99 IV/99 I/00 II/00 III/00 IV/00 I/01 II/01 III/01 IV/01 I/02 II/02 III/02 IV/02 I/03
Quarter of year
Number of prescriptions
Macrolides Cephalosporins Amoxicillin Doxycycline Quinolones Amoxicillin with clavulanic acid Penicillin-V Trimethoprim Sulfa-trimethoprim combinations
The more ill the patient is with his/her respiratory tract infection, the more probable it is that the causative agent is Streptococcus pneumoniae. At this level of macrolide resistance it is probably safer in such cases to treat the patient with penicillin or amoxicillin, (or cephalosporin or doxycycline) than with macrolides. The usefulness of beta-lactams in such cases is also supported by recent meta-analyses of hospitalized patients with community acquired pneumonia, which conclude that empirical coverage of atypical pathogens in primary antibiotic choices does not give benefit in terms of survival or clinical efficacy over beta-lactam antibiotics (Mills et al., 2005, Shefet et al., 2005).
First-generation cephalosporins are recommended in Finland as the drug of choice for skin infections, and in the case of penicillin allergy, for tonsillitis.
These diagnoses explained 42% of the consumption in 1994, 59% in 1998 and 65% in 2002. First-generation cephalosporins seldom result in any microbiological benefit compared with firstly recommended drugs in sinusitis, bronchitis or pharyngitis. The proportion of second-generation cephalosporins is low, which is correct as they are recommended only as the second or third option, even though they can be microbiologically justified in some cases of otitis media and sinusitis. Furthermore, it is noteworthy that although cefuroxime axetil represented only 7% (38/516) of all cephalosporins used in the Pirkanmaa study, it accounted for 85% (17/20) of all reported severe cases of cephalosporin-induced colitis in Finland during 1990–1995 (Lumio, 1996). Using such medicine to treat common, mild or self-limiting respiratory infections cannot be justified when safer alternatives are available.
94
The proportion of macrolides among all antibiotics was only slightly rising during the MIKSTRA study, which is favourable. According to sales statistics its use seems to have begun to decline nationally since then. However, while in the Pirkanmaa study macrolides were mainly used for lower respiratory tract infections, bronchitis (44%) and pneumonia (14%), in the 1998 MIKSTRA data pneumonia comprised only 5% and other LRTIs 31% of all indications for use of macrolides and half of them were used to treat otitis media or sinusitis. The situation has remained similar in the following years of MIKSTRA. From that point of view, development has proceeded in an undesirable direction as pneumonia is still the main indication for use of macrolides and it is recommended only as a secondary option for upper respiratory tract infections.
Use of macrolides could, however, be rather easily reduced by some 45–50%.
Over eighty per cent of all macrolides in the MIKSTRA study were used to treat otitis media (29%), acute bronchitis (24%), sinusitis (21%) and unspecified URTI (9%) (Figure 17, page 77). The proportion of macrolides in the treatment of otitis media was 16% and in sinusitis 14% and this comprised together half of all macrolide use. At least one third of this could easily be replaced with other antibiotics, primarily with penicillin group. Another third of patients could probably avoid antibiotic all-together by improved diagnostics (Blomgren et al., 2004, Mäkelä & Leinonen, 1996, Palmu et al., 1999) or by a follow-up for one to three days from the onset of symptoms in the case of acute otitis media (Jensen
& Lous, 1998, Marchetti et al., 2005, McCormick et al., 2005) and a week in the case of suspected sinusitis (de Bock et al., 2001, Gwaltney Jr, 2005c). This would mean already some 25–30% reduction in the total consumption of macrolides. In the case of acute bronchitis and unspecified URTI, which comprise one third of all macrolide use, the reduction could be even grater. A conservative estimate suggests that one third of these patients could avoid any antibiotic use and another third could be treated with penicillins or doxycycline instead of macrolides leading to approximately 20% additional reduction in the total consumption of macrolides.
Sulfa and sulfa-trimethoprim combinations are regarded as less desirable drugs in many countries because of the severe side-effects (Anonymous, 1995, Friis, 1987) or resistance (Kristinsson, 1999). It seems that sulfa-trimethoprim combinations are not used to treat respiratory tract infections in many other countries to the same extent as they are in Finland and Iceland. Sulfa-trimethoprim mixtures have a relatively pleasant taste to children, they are inexpensive, and often microbiologically justified. Severe side effects of sulfa-trimethoprim combinations have not been reported among children in Finland. A reasonable use of this antimicrobial combination for treating otitis media is not contradictory to the Finnish recommendations, but it is not recommended as a primary choice. Observations made in Iceland on the colonization of multiresistant strains of pneumococci in children after recurrent treatment with antibacterials, especially sulfa-trimethoprim combinations (Arason et al., 1996) warrants, however, cautiousness and close observation of the development of resistance also in our country. The declining use of this antimicrobial
95 combination is rather satisfying than worrying as it is not a recommended choice in uncomplicated UTIs either.
A high proportion of tetracyclines in total antibiotic use is not generally regarded as reflecting high quality prescribing at a European level because of the high level of resistance. In our data doxycycline was practically the only tetracycline prescribed in health centres. However, in the national sales data, roughly one third of all tetracyclines are other than doxycycline, but obviously they are prescribed by medical professionals other than GPs. Doxycycline can be regarded as even a recommendable choice in Finland in treating many respiratory tract infection in adults if the penicillin group cannot be used or if there is a suspicion of infections caused by Mycoplasma or Chlamydia pneumoniae (Korppi et al., 2003). The resistance of most respiratory pathogens to doxycycline is fairly low in Finland in contrast to already worrying high resistance levels to macrolides as well as sulfa-trimethoprim combinations, which are commonly used alternatives in these cases. It is possible that tetracyclines are used differently in Finland in respect to dosing and/or indication than in other countries having a high level of resistance.
Commercial information is an important modifier of the drug prescribing of practising doctors. Although the correctness of the information material is supervised by the authorities and it is usually not false as such, marketing may be unbalanced as a result of commercial interests and thereby direct prescribing practices in an inappropriate direction. A new product is launched by active marketing and competitors also need to respond to it. It is more profitable to put marketing efforts on new, more expensive drugs than old, less expensive alternatives. Commercial information is also product centred and does not, in most cases, deal with the patient's problem comprehensively. A probable explanation of the unnecessary increase in the use of amoxicillin with clavulanic acid is that it is marketed more actively than amoxicillin alone, which gives a false impression on the resistance situation and the status of this drug among other options. Penicillin-V again, is not a profitable substance from the point of view of pharmaceutical marketing, which leads to an unnecessary invisibility.
Marketing of new macrolides for infections of the upper respiratory tract has been active and evidently successful from the industry's point of view.