Medical devices in Sweden : Industrial structure, production and foreign trade 1985-2002

Abstract: This licentiate thesis uses descriptive, mainly official,Swedish statistics to analyse industrial structure, productionand foreign trade in an industry that is traditionallydifficult to describe in numbers, that of medical devices. Forthe purposes of the thesis, the Swedish Medical Device industryis defined as companies classified in the SE-SIC manufacturingcodes 33101 (medical equipment and instruments, etc), 33102(dental products) and 35430 (invalid vehicles). Also otherbranches contribute, notably parts of SIC 51460 (wholesale inmedical equipment and pharmaceutical goods) and 73103 (medicalresearch and development) although their medical device volumecannot be specified. Additional items have been identified interms of specific product groups rather than as "belonging" toa specific SIC industrial code.Taken together, this is considered to correspond reasonablywell to the scope of the field as defined by the Global MedicalDevice Nomenclature (GMDN), a new European standard forclassifying medical devices in a more generic way than do theEuropean Medical Device Directives (or other pieces oflegislation). No quantification according to GMDN can be madeas yet, however, as that requires changing reporting habits inindustry as well as in official statistical classification andnomenclature regimes.With the manufacturing code SE-SIC 33101 as main object, thestudy for the first time presents data on the regionaldistribution, size classes of employment, company starting timeand company dynamics, in the form of entries to and exits fromthe code, over a six-year period. The latter analysis includesa follow-up of the "exits", some firms reappearing in otherparts of industry and others disappearing–surprisinglyfew among them being limited companies. Although this industryis comparatively mature, considerable mobility among themid-sized companies is indicated for reasons of real changes or(to some degree) factors inherent in the industrialclassification system. Some structural changes in companies inthe ≥50 employees bracket are identified. It is notedthat American actors, directly or indirectly, are increasinglyinvolved with the medical device industry in Sweden, and that anumber of technology-based companies that were started mostlyin the early eighties have recently reached the 50+ employeelevel.The analysis of identifiable production and internationaltrade in medical devices spans a period of 17 years based onofficial statistics following the HS/CN nomenclatures. Adatabase has been built, bottom-up, from the 8-digit CN levelwith production, exports and imports values for close to 100items collected in 12 product groups, for presentation purposesgrouped under three main headings. Compound annual growth ratesfor the latter are presented for three five-year periods1985-2000, showing that Swedish production and exports have hadan overall growth of 10 per cent p.a. This has kept Swedenahead of the international overall growth of 6- 7 per cent p.a.in recent years, products in the main group "Aids&Implants" growing more than 20 per cent p.a. Growth rates inthe most recent five-year period are lower, however. Healthynet exports figures are presented, the figure for 2002nominally representing 40 per cent of the production value incurrent as well as constant prices.Production figures are given at industry (local unit) levelas well as at product group level. The product-based figuresidentified for Production 2001 are estimated to SEK 13,3billion, Exports to SEK 13,7 billion and Imports to SEK 9,7billion. Figures for the Apparent Domestic Market arecalculated for the corresponding entities. It is obvious,however, that the statistics do not capture the real productionvalue as exports exceed production both at overall level and inmajor product groups, particularly those on a high systemstechnology level. The situation is not uncommon for a number ofreasons; further, cases in the statistics methodologyliterature confirm that medical instrument-related codes areliable to this phenomenon. Corrections, including adjustmentsof both production and exports values, are possible butdemanding already at one individual 4-digit HS/CN level. This,therefore, must be considered outside the scope of an academicstudy.The basic tablework developed for this thesis will be madefreely available to external parties for their own use providedthe author, with contact details, is named as the source.(Processing for commercial purposes is not expected, however.)Any suggestions for improvements are welcomed.