An 11-year total population study of Western Australia found the prevalence of babies born with upper limb anomalies to be 19.76 in 10,000 live births [3]. Forty-six percent of those affected had another non-hand congenital anomaly. Fifty-one percent had bilateral hand anomalies, and 17 % had multiple different hand anomalies. The most common anomalies were failures of differentiation (35 %), duplications (33 %), and failures of formation (15 %). Congenital upper limb anomalies were more common in boys; preterm, post-term, and multiple births; and older mothers. No significant differences in prevalence or frequency of anomalies were found between whites and nonwhites, left and right sides, and in babies that survived and those who died shortly after birth.

Similarly, an 11-year total population study of the Stockholm region of Sweden found a recorded incidence of congenital anomalies of the upper limb of 21.5 per 10,000 live births [4]. Fifty-four percent of the children with congenital anomalies of the upper limb were boys. The anomalies affected the right side only in 30 %, the left side only in 33 %, and both sides in 37 %. Non-hand anomalies were recorded in 23 % of the children with congenital anomalies of the upper limb, most commonly in the lower limbs. In 17 % of the affected children, there was a known occurrence among relatives. Failure of differentiation was the most common category (47 %) followed by duplication (26 %), failure of formation (18 %), undergrowth (3 %), generalized abnormalities and syndromes (2.4 %), overgrowth (1.7 %), and constriction ring syndrome (1.5 %).

There are more total population studies of limb deficiency anomalies, for example: a 9-year total population study of the national incidence of upper limb deficiencies in Finland found an incidence of congenital deficiency anomalies of the upper limb of 5.26 per 10,000 live births [5]. These studies approximate the “failure of formation” category of complete CAUE population studies (Table 2.2).

Incidence figures derived by extrapolation from surveys of patients presenting for treatment show slightly lower incidence: an estimated 16–18 per 10,000 births [9–11]. It is thought that these population studies may underestimate incidence, as the milder deformities may never present for treatment. A comparison of a population-based study and clinic registry of Swedish children with CAUE showed an underestimation of incidence by 6 % in the clinic registry, and a low degree of correlation of classification of anomalies [12].

The IFSSH classification is a useful tool for classifying most CAUE and enables comparison between studies, but is based on theories of embryological failure and is subject to some differences of interpretation. Ambiguities in the categorization of anomalies may then lead to differences of incidence of certain classifications [13]. For instance, the IFSSH classification could classify polydactyly with complex syndactylies as duplication, but for clinical purposes it fits better into the category of failure of differentiation. Miura et al. [14] and Ogino [15, 16] suggested that a common teratological mechanism causes cleft hand, syndactyly, and polydactyly and that they should be put into a new category: failure of induction of digital rays. Classifying congenital absence of digits is also ambiguous; the distinctions between brachysyndactyly, symbrachydactyly (atypical cleft hand), and transverse arrest are not clearly defined.

In the Stockholm study thumb hypoplasia was categorized as failure of formation, longitudinal arrest, and radial ray deficiency, whereas in the study from Western Australia, thumb hypoplasia was categorized as undergrowth. The Stockholm study showed a much lower frequency of undergrowth as a result. There was also a surprisingly large disparity between the categories of transverse arrest and symbrachydactyly regarding associated non-hand anomalies. Other differences in relative frequencies are also likely caused by other differences of interpretation of classification strategies.

Epidemiologic studies are important for health care planning, detecting changes in incidence over time, and comparing differences among regions. These two total population studies of CAUE agree on total incidence figures. These population studies are slightly higher than the estimated 0.16–0.18 % incidence for CAUE of in surveys of patients presenting for treatment (Table 2.3) [9, 10, 17–19]. It is assumed that this is due to the fact that milder deformities may never present for treatment.

These studies do, however, reveal the difficulties in comparing studies owing to different classification strategies and weaknesses within the IFSSH classification. For example, two studies of CAUE in Edinburgh, UK [2, 10] and two studies from Japan [17, 18] show markedly different relative frequency of incidence of duplication; presumably such a finding in ethnically similar populations is due to differences in classification (see Table 2.3). We hope that the ongoing discussion of classification systems for CAUE (see Chap. 1) will inspire improvements in registration and population studies.