Screening of Newborns and Infants for Developmental Dysplasia of the Hip: A Systematic Review

Abstract

Background

Developmental dysplasia of hip (DDH) is a common disorder of childhood and has a good prognosis when treated at an early age. In spite of being a significant concern, many children with DDH are not picked early and present late at walking age. In our country, it is presumed to be due to absence of a national policy for screening of DDH. Screening programmes including the combination of clinical and radiological methods in different ways have been suggested. However, the exact method of screening is controversial.

Purpose

To analyze effectiveness and cost-effectiveness of various screening methods for DDH.

Study Design

Systematic review.

Methods

This review was conducted in accordance with PRISMA guidelines. Medline database was explored for original case series and randomized clinical trials. Inclusion criteria were English language, screening for DDH in neonates, sample size more than 500, and studies with minimum duration of one year.

Results

Thirty-four studies were selected to write the manuscript. This included 23 studies looking for effectiveness of a screening programme and 11 studies comparing various outcomes of different screening strategies. A trend favoring universal ultrasound screening was observed

Conclusion

The literature supports universal ultrasound screening and has proved its cost-effectiveness. However, considering the logistic and financial challenges in our country, immediate implementation of universal ultrasound screening seems impractical. In the absence of any current guidelines for screening for DDH in India, we suggest professional organizations involved in the care of children and public health policy-makers to come together to develop national screening guidelines for DDH.

Introduction

Developmental dysplasia of hip (DDH) is a common disorder of childhood with global incidence of 1.6–28.5 per 1000 live births [1]. It refers to a spectrum of hip problems ranging from stable hip with mild acetabular dysplasia to more severe forms with frank dislocation of hip joint. When detected early, it can be treated effectively with simpler means which are readily available and acceptable to the family and owns good prognosis [2]. Treatment of late-presenting DDH is complex and needs surgical intervention. This increases morbidity, cost of treatment, and overall burden on healthcare [2, 3]. Previous studies have shown an incidence of up to 9.2 per 1000 live births for DDH in India [4, 5]. Looking at the huge population of our country, a large number of cases can be anticipated. In spite of being a significant concern for paediatricians and orthopaedicians, many children with DDH are not picked early and present late at walking age [6]. In India, this can be attributed to absence of a national policy for screening of DDH as clinicians do not feel it obligatory to screen every newborn. Furthermore, the ideal screening method for DDH is yet to be sought. Screening programmes including combination of clinical and radiological methods in different ways have been suggested and are being used in different countries [7, 8]. However, exact method of screening is controversial. Through this study, an attempt was made to analyze effectiveness of various screening methods for DDH in decreasing the rate of late diagnosed DDH (LDD) and their cost-effectiveness was studied.

Materials and Methods

This systematic review was performed as per PRISMA (Preferred Reporting Items for Systematic reviews and Meta-Analyses) guidelines [9]. To avoid ambiguity as much as possible, inclusion and exclusion criteria for selection of studies were decided before initiating literature search. The Medline database was explored using combination of MeSh terms; DDH, newborn, clinical screening, and ultrasonography using PubMed, EMBASE, and Google Scholar. Only original case series and randomized clinical trials were included. Related reviews were searched for additional references. Inclusion criteria were English language, screening for DDH in neonates (age < 4 weeks), sample size more than 500, and studies with minimum duration of 1 year. This included studies looking for diagnostic effectiveness of a screening method for DDH in newborns and those comparing one screening method with another in terms of outcome. Case reports, reviews, letter to the editor, and unpublished data were excluded from final list of studies. Attempt was made to eliminate bias in selecting studies through factors based on reliability, internal consistency, and empirical evidence if available. The original studies with available full-length articles were collected. Bibliography of all eligible studies was explored further to find additional studies. Any difference in opinion between the authors was discussed and resolved. Selected studies were assessed for quality using modified Jadad Scale [10] for randomized-controlled trial and GRADE [11] (Grading of Recommendations, Assessment, Development and Evaluation) for case series and those with better scientific evidence were selected to write the manuscript. The selected effectiveness studies were reviewed for method of screening and its effectiveness in decreasing incidence of LDD. Comparative studies were explored for various outcomes of different screening methods like sensitivity, specificity, treatment rate, mean age at treatment, mean age at first operation, cost-effectiveness of each method, and challenges encountered, in implementation of screening strategy.

Results

Two thousand nine hundred seventy-seven studies were screened initially after online search and manual search of bibliography of individual articles (Fig. 1). After further screening, 375 studies on screening for DDH in neonatal age were identified. 106 studies met inclusion criteria, and after quality assessment, 34 were selected to write the manuscript (Tables 1, 2, 3 and 4) [6, 7, 12–43]. This included 23 studies looking for effectiveness of a screening programme and 11 studies comparing various outcomes of two or more screening strategies. Current review covers around 67 lakh neonates screened by one or more methods and the largest series was by Broadhurst et al. [6]. Duration of studies ranged between 1 and 40 years (mean 10.32 years) and longest study was by Myers et al. [7]. Different age criteria were used for definition of late presentation. Majority considered cases presenting after 3 months as late presentation [13–16, 20–23, 34]. Duppe et al. [37] used the shortest cut off mark of 1 week, whereas Broadhurst et al. [6], Milligan et al. [24], and Donnelly et al. [25] used the longest period of 12 months.

Fig. 1.

Flowchart of literature search as per PRISMA guidelines

Table 1.

Studies evaluating effectiveness of clinical screening alone

Sr no	Authors (year)	Sample size (N)	Method of screening	Duration of study (years)	Rate of LDD (per 1000 live births) (definition of LDD)	Conclusion
1	Lehmann et al. 1981 [12]	176,000	UC at birth by orthopaedic surgeon vs orthopaedic resident	5	Surgeon 0.3 Resident 0.8 (> 4 weeks)	Incidence of LDD decreased and was further less in experienced hands. Recommended UC screening
2	Wenger et al. 2019 [13]	1,013,589	UC in neonates by paediatrician. At 6–8 weeks, 6 months and 10–12 months by GP	18	0.12 (> 2 weeks)	Incidence of LDD, age at diagnosis and disease severity decreased. Recommends serial clinical hip examination after initial screening at birth
3	Studer et al. 2016 [14]	> 100,000	UC in neonates by an experienced health practitioner. At 6–8 weeks and 6–9 months by GP or paediatrician	11	0.77 (> 3 months)	Incidence of LDD increased despite the screening programme. Advised for increased awareness, education and to avoid inappropriate swaddling
4	Myers et al. 2009 [7]	41,563	UC within a week of delivery by an orthopaedic surgeon	4	0.1 (> 6 months)	UC screening by experienced hand significantly reduces the incidence of late-presenting (walking) DDH

UC universal clinical, LDD late diagnosed DDH, GP general practitioner

Table 2.

Studies evaluating effectiveness of universal clinical and selective ultrasound screening

Sr no.	Authors (year)	Sample size (N)	Method of screening	Duration of study (years)	Rate of LDD (per 1000 live births) (definition of LDD)	Conclusion
1	Broadhurst et al. 2019 [6]	3,635,163	UC within 72 h and S-USG at 2/6–8 weeks. Further clinical assessment at 6–8 weeks by GP	26	1.28 (> 12 months)	Incidence of LDD did not reduce. Recommends further research and debate for use of U-USG screening
2	Talbot et al. 2017 [15]	64,670	UC at birth and 6–8 weeks and S-USG at 6–8 weeks	15	0.28 (> 3 months)	LDD still occurs. Recommends for update of screening programme, review training of physicians and further assessment after 6 week check
3	Clarke et al. 2012 [16]	107,440	UC within 72 h and S-USG at 2/6 weeks. Further clinical assessment at 6 weeks by GP	20	0.34 (> 3 months)	Recommends UC and S-USG screening, one stop treatment and further clinical monitoring till 12 months
4	Geertsema et al. 2019 [17]	3536	UC at 2–3 weeks and S-USG at 6 weeks/3 months	4	Persistent DDH at 18 months 0.3%	Delayed S-USG screening at 3 months in children with clinically stable hip but having risk factors for DDH is safe and this approach could prevent overtreatment
5	Laborie et al. 2014 [18]	81,564	UC and S-USG at 1–3 days and further clinical assessment at 6 weeks, 6 months and 1 year	16	0.32 (> 4 weeks)	Recommends S-USG screening as it gives acceptable rates of early treatment and low rates of LDD
6	Paton et al. 2002 [19]	28,676	UC within 48 h and S-USG at 2/6 weeks. Further clinical assessment at 6 weeks by GP	8	0.87 (> ?)	Recommended against national S-USG screening as it did not reduce the overall rate of LDD
7	Lisle et al. 2011 [20]	30,000	UC at birth, 6 weeks, 3, 6 and 12 months. S-USG at 6 weeks	1	0.57 (> 3 months)	Incidence of LDD increased three times the previously established rate. Suggested further research
8	Chan et al. 1999 [21]	118,379	UC at birth and 6 weeks. S-USG at 6 weeks. Further clinical check till 2 ½ years	5	0.19 (> 3 months)	Recommends UC and S-USG screening with further clinical assessment in later life
9	Sahin et al. 2004 [22]	5798	UC at birth. S-USG at 3–8 weeks. Further clinical assessment at 2, 4, 6, 9 and 18 months	7	1.37 (> 3 months)	Recommends S-USG screening and serial clinical hip examinations till walking age
10	Boeree et al. 1994 [23]	26,952	UC at birth and S-USG at 2/4–6 weeks and clinical assessment at 6 weeks	4	0.22 (> 3 months)	Delayed S-USG screening is more effective than UC screening alone
11	Milligan et al. 2020 [24]	170,580	UC and S-USG within 6 weeks and further clinical assessment at 4 months	6	0.30 (> 12 months)	Continuous monitoring of a S-USG screening programme improves the outcome
12	Donnelly et al. 2015 [25]	75,856	UC at birth, 6–8 weeks and 4 month combined with S-USG within 10 days or at 6 weeks	2	0.42 (> 12 months)	Recommends additional clinical assessment at 4 months in addition to UC at birth and S-USG
13	Choudry et al. 2018 [26]	28,241	UC within 72 h + S-USG at 2/6–8 weeks. Further clinical assessment at 6–8 weeks by GP	4	– (> 6 weeks)	A decrease in positive predictive value (PPV) of screening and increase in rate of surgery was observed

UC universal clinical, S-USG selective ultrasonography, LDD late diagnosed DDH, GP general practitioner

Table 3.

Studies evaluating effectiveness of universal ultrasound screening

Sr no.	Authors (year)	Sample size (N)	Method of screening	Duration of study (years)	Rate of LDD (per 1000 live births) (definition of LDD)	Conclusion
1	Thallinger et al. 2014 [27]	> 80,000	UC + U-USG at birth and 6–8 weeks	16	–	U-USG screening produces a distinct and progressive decrease in number of hip surgeries
2	Ulziibat et al. 2020 [28]	176,388	U-USG at 1–2 days after birth	2	–	U-USG screening provides high screening, quality surveillance and appropriate treatment rate. It can be deployed successfully in developing countries
3	Biedermann et al. 2018 [29]	28,092	U-USG at birth and UC + U-USG at 6–8 weeks	5	00	U-USG screening reduced rate of secondary DDH surgery later in life. Recommends early U-USG screening
4	Von kries et al. 2003 [30]	50,000	UC and U-USG before 6 weeks	5	0.14 (> 9 weeks)	Recommends early U-USG screening and suggests for training of doctors in doing USG
5	Marks et al. 1994 [31]	14,050	U-USG	3	00	Recommends U-USG screening
6	Olsen et al. 2018 [32]	4245	UC + U-USG within 3 days	8	0.5 (> 4 weeks)	Adding U-USG to UC screening doubled the treatment rate without a significant reduction in incidence of LDD

UC universal clinical, U-USG universal ultrasonography, LDD late diagnosed DDH

Table 4.

Comparative analysis of different screening strategies

Sr no	Authors	Sample size (N)	Methods compared	Duration of study (years)	Definition of LDD	NS	UC	UC + S-USG	UC + U-USG	Outcome
1	Tredwell et al. 1990 [33]	–	NS UC	–	–	TC 19,101.75$/1000 live births	TC 3391.74$/1000 live births	–	–	UC is more cost-effective and has economic benefit of 15,143.71$/1000 live births
2	Rosendahl et al. 1994 [34]	11,925	UC UC + S-USG UC + U-USG (1st week)	2	> 3 months	–	TR 18 LDD 1.3	TR 20 LDD 0.7	TR 34 LDD 0.3	Effect of USG screening in reducing incidence of LDD was not statistically significant
3	Geitung et al. 1996 [35]	–	UC UC + U-USG	–	–	–	TC 17,205,620 NOK	–	TC 16,900,000 NOK	UC + U-USG is not cost-effective. However, U-USG is cost-effective when UC is omitted. Recommends selective screening
4	Clegg et al. 1999 [36]	–	UC UC + S-USG UC + U-USG	20	–	–	SI 6.5/year AFO 12.4 MSC 5110£/1000 live births	SI 5.4/year AFO 14.2 MSC 3811£/1000 live births	SI 2.5/year AFO 6.7 MSC 468£ / 1000 live births	Overall cost for management of DDH was comparable. Recommends universal USG screening
5	Duppe et al. 2002 [37]	55,267	UC and UC + S-USG (1st week)	20	> 1 week	–	LDD 0.51 TR 19.9 RR 31.7	LDD 0.07 TR 6.6 RR 12.9	–	Recommends UC + S-USG
6	Holen et al. 2002 [38]	15,529	UC + S-USG UC + U-USG (1st week)	5	> 4 weeks	–	–	TR 8.6 LDD 0.65	TR 9.6 LDD 0.13	Recommends S-USG as difference was not statistically significant
7	Dezateux et al. 2003 [39]	400,000	NS UC UC + S-USG UC + U-USG	–	–	TR 0.12% SR 0% SI 0.12% FTO 75%	PSR 2.1% DR 35% TP 0.04% FP 2.08% TN 97.79% FN 0.078% TR 0.52% SR 0.42% SI 0.10% FTO 78%	PSR 8.1% DR 60% TP 0.07% FP 8.05% TN 91.82% FN 0.04% TR 0.75% SR 0.70% SI 0.05% FTO 88%	PSR 7.7% DR 76% TP 0.09% FP 7.57% TN 90.31% FN 0.02% TR 0.54% SR 0.51% SI 0.03% FTO 92%	Clinical screening alone is only marginally beneficial when compared to no screening. USG-based screenings are more effective
8	Roovers et al. 2005 [40]	7236	UC UC + U-USG (1, 2, 3 month)	14 months	> 3 months	–	TR 27 RR 192 LDD 8 S 76.4%	–	TR 46 RR 76 LDD 6 S 88.5%	U-USG detects more cases of DDH and at an early age. However, it leads to overtreatment. Recommends delayed U-USG at 2–3 months
9	Gray et al. 2005 [41]	629	UC UC + U-USG	3	–	–	SR 48% SI 8% TC 1488$ / patient	–	SR 37% SI 7% TC 1298$ / patient	U-USG does not increase cost burden on health services and family and decreases SR & SI. Recommends UC + U-USG
10	Thaler et al. 2011 [42]	83,016	UC And UC + U-USG (1st week)	8	–	–	CR 25.2/year SI 17.8/year TC 410,000 €/year	–	CR 7.0/year SI 2.6/year TC 467,000 €/year	Number of interventions decreased by 75.9% since introduction of U-USG and it is cost-effective. Recommends UC + U-USG
11	Tschauner et al. 2011 [43]	–	UC UC + U-USG (< 6 weeks)	30	–	–	AIT 5.5 CR 88.7% OR 11.3%	–	AIT 2 CR 98.9% OR 1.1%	U-USG-based treatment is safer, shorter, and simpler. Recommends UC + U-USG

NS no screening, UC universal clinical, S-USG selective ultrasonography, U-USG universal ultrasonography, LDD Late Diagnosed DDH (per 1000 live births), TR treatment rate (per 1000 live births), RR referral rate (per 1000 live births), CR close reduction, OR open reduction, SI surgical intervention, TC total cost, PSR positive screening result, DR detection rate, TP true positive, FP false positive, TN true negative, FN false negative, SR splintage rate, FTO favourable treatment outcome, S sensitivity, AIT age at initial treatment (months), AFO age at first operation (months), MSC mean surgical cost

Screening methods discussed for their effectiveness were either universal clinical alone or combination of universal clinical with selective or universal ultrasonography (USG). Four studies observed outcomes of universal clinical neonatal screening alone and analyzed its effectiveness in reducing the incidence LDD (Table 1) [7, 12–14]. Three of them found it effective when compared to earlier data in their respective area before implementing the screening programme [7, 12, 13]. They observed that incidence was further less when screening was done by an experienced physician and when serial clinical hip examinations were done. It also effectively reduced the mean age at diagnosis of DDH and disease severity. However, Studer et al. [14] reported increased incidence of LDD despite universal clinical screening in neonates and repeated hip examination at 6–8 weeks and 6–9 months. They advised for more education and awareness among healthcare workers and for avoiding inappropriate swaddling of infants to reduce incidence of late onset DDH. Mean incidence of LDD after universal clinical neonatal screening alone was 0.37 / 1000 live births (range 0.1–0.77/1000 live births).

Thirteen studies analyzed effectiveness of universal clinical neonatal screening combined with selective USG screening (Table 2) [6, 15–26]. Clinical examination of hip was performed unanimously in the first week of life by all. However, there was discrepancy in timing of doing the USG evaluation. Laborie et al. [18] and Olsen et al. [32] performed the USG scan early in first week, while Geertsema et al. [17] did the same at 3 months. However, majority scanned hip between 3 and 8 weeks of life [15, 20, 22, 24]. Five studies followed a differential strategy and scanned the unstable hips early within 2 weeks, whereas those with normal clinical screening but with risk factors for DDH were scanned at 6 weeks [6, 16, 23, 25, 26]. Eleven out of 13 studies maintained further check by repeat clinical hip examination till age between 6 weeks and 30 months [6, 15–18, 21–26]. This strategy decreased the incidence of LDD when compared with single hip examination; however, none of these were successful in completely eliminating it. Satisfactory outcome of a selective screening programme was reported by eight studies [16–18, 21–25], whereas five suggested further research and debate regarding introduction of universal USG screening [6, 15, 19, 20, 26]. The mean incidence of LDD after universal clinical and selective USG screening was 0.53/1000 live births (range 0.19–1.37/1000 live births).

Six studies reported outcome of universal USG screening of hip for DDH and five observed satisfactory results (Table 3) [27–32]. Two of them claimed that it eliminated incidence of LDD after following newborns till 5 years of age [29, 31]. The incidence of LDD was not mentioned specifically in two studies, but they recommended this method after observing a distinct and progressive decrease in number of hip surgeries and an acceptable treatment rate [27, 28]. They further suggested for training of doctors in doing USG scans. Contradictory to this, Olsen et al. [32] reported significant increase in treatment rate without any significant reduction in incidence of LDD. Mean incidence of LDD after universal USG screening was 0.13/1000 live births (range 00–0.5/1000 live births).

Comparative analysis of different screening strategies was done by 11 studies (Table 4) [33–43]. This included their sensitivity, specificity, overall referral rate, overall treatment rate, overall splintage rate, number of surgical interventions, mean age at initial treatment, favourable treatment outcome, mean age at first operation, incidence of LDD, and cost-effectiveness. Tredwell et al. [33] reported that universal clinical screening alone was more cost-effective when compared to no screening for DDH. Dezateux et al. [39] observed only marginal benefit of isolated clinical screening and found the USG-based screenings more effective. However, a preference between selective and universal USG screening was not made as both methods had their own limitations. For selective USG screening, this included uncertainty about indications of doing scan and ultrasound indications for splint therapy. For universal USG screening, this included high false-positive rate and unnecessary treatment of normal children with consequent increase in complications of abduction splinting like avascular necrosis, femoral nerve palsy, pressure sores, and parental anxiety. Similar results about universal USG screening were observed by Roovers et al. [40] and they suggested for delayed screening at 2–3 months to avoid overtreatment. Overall, universal USG screening showed increase in treatment rate but a better detection rate and favourable treatment outcome (Fig. 2). At the end, six studies reported universal USG to be cost-effective, and supported combination of universal clinical and universal USG screening [33, 36, 40–43]. Two studies suggested against it and recommended for combination of universal clinical and selective USG screening [34, 35, 38].

Fig. 2.

Comparison of outcomes of different screening methods. TR treatment rate, DR detection rate, FTO favourable treatment outcome, UC universal clinical, S-USG selective ultrasonography, U-USG universal ultrasonography

Discussion

Screening of newborns for hip instability is being practiced for several decades with an aim to reduce incidence of late-presenting DDH [7, 8]. However, there is lack of consensus about the ideal method. An ideal screening method should be easy to use, readily available, cost-effective and acceptable to the patient. It should also have adequate sensitivity and specificity to detect the problem. Broadly, screening for DDH involves use of clinical examination of hip, radiological investigations, or combination of both (Table 5). Till now, none of the programmes have been successful in eliminating incidence of late-presenting DDH. Shorter et al. [44] in his Cochrane review, did not find any conclusive evidence to recommend a particular method of screening.

Table 5.

Various screening methods for DDH and their respective mean incidence of late diagnosed DDH

Method of screening	Description	Method used	Mean incidence of late diagnosed DDH (per 1000 live births)
UC	Clinical hip examination of all newborns for signs of instability and risk factors of DDH	Barlow test Ortolani test	0.37
UC + S-USG	Clinical hip examination of all newborns USG of hip for those with positive clinical screening USG of hip for those with negative hip screening but having one or more risk factors of DDH and those with an inconclusive clinical screening	Clinical: Barlow test Ortolani sign USG: Graf’s method	0.53
UC + U-USG	Clinical hip examination of all newborns USG of hip for all newborns	Clinical: Barlow test Ortolani sign USG: Graf’s method	0.13

In absence of a national screening programme, risk factor-based screening is resorted to by individual practitioners [45, 46]. Newborns with one or more risk factors for DDH are examined clinically and/or sonographically to confirm hip instability. The knowledge of these risk factors and awareness of possible hip abnormality is a prerequisite to avoid missing DDH. Major known risk factors are family history of DDH and breech presentation. The minor risk factors include first born child, female gender, oligohydramnios, prematurity, twins, excess birth weight, torticollis, and post-natal swaddling [14, 47]. An increased risk of DDH was also observed with idiopathic clubfoot deformity and high altitude [48, 49]. Sixty percent of overall dislocations have been reported to occur in children with one or more risk factors. Furthermore, the risk of DDH increases with increased number of risk factors [50]. However, mere presence of a risk factor does not necessarily indicate presence of DDH. Paton et al. [51] reported much less incidence of dislocation (1 in 75) in infants with risk factor when compared to those with clinical instability (1 in 11). Similarly, DDH is often detected in children without any risk factor. Thus, screening for risk factors alone is not recommended as an ideal method of screening for DDH.

Effectiveness of universal clinical neonatal screening in reducing incidence of late diagnosis of DDH was reported by three out of four studies in current review [7, 12, 13]. Moreover, decrease in age at detection and disease severity was observed [13]. Commonly used clinical tests were Barlow test, Ortolani sign, limited hip abduction, asymmetric thigh folds, and limb length discrepancy. A combination of Ortolani and Barlow test with specificity of 95% is the most commonly used method [12, 13]. They are easy to perform and lack cost issues. However, they are less sensitive (28%) and are unable to pick up irreducible dislocation [13]. Similarly, subluxable hip and dysplasia without dislocation are not detected and may progress and present later [7]. The signs can be very subtle in a newborn posing challenge even to an experienced paediatric orthopaedic surgeon [52]. Other clinical tests may indicate presence of DDH, yet are not suitable as a routine screening test due to poor sensitivity and specificity [53, 54]. Effectiveness of a clinical screening programme alone in reducing the incidence of late presentation of DDH has been refuted by many studies [14]. Even increased incidence in spite of an ongoing clinical neonatal screening programme was observed [14]. However, effectiveness can be improved significantly if clinical screening is done by experienced physicians [12, 15]. The newborn physical examination standards and competencies programme was launched in UK aiming to ensure that all physicians involved in screening programme have adequate training and competency to do that [55]. It is also aimed to assess and audit their performance periodically. Where experienced physicians are not available, paramedics can be involved in the programme after adequate training without compromising the result [56].

DDH includes both morphological abnormality and instability of hip. Unlike an unstable hip, morphological errors are missed by clinical examination. Application of USG for diagnosis of congenital hip dislocation was first reported by Graf in 1980 [57]. Since then, use of USG has decreased the incidence of late presentation and overall, a fall in surgical rates [18, 23, 30]. It is more sensitive and specific when compared to clinical tests and picks up both morphological abnormality as well as instability (Table 6). It is safe, non-invasive and is the preferred radiological tool for screening DDH in children below 6 months of age when femoral head is still cartilaginous. Use of USG can be ‘universal’ or ‘selective’. There is lack of consensus about which method to prefer, definition of an abnormal hip, and importance of visible morphological abnormalities in USG. Controversy also persists regarding the timing of ultrasound screening [12, 15, 17, 20]. Graf et al. suggested USG screening of all newborns in first week of life [58]. Others favour late screening after 6 weeks as most unstable hips detected at birth become stable spontaneously by 4 weeks [23, 31]. In current review, six studies were in favour of early USG screening in first week, whereas seven studies supported screening after 3 weeks [15, 17, 20, 22, 24]. Five studies followed a differential strategy and scanned clinically unstable hips within 2 weeks, whereas delayed screening at 6–8 weeks was done for those with normal clinical screening but with risk factors for DDH [6,16,23,25,26]. USG screening is further criticized for having poor inter-observer and intra-observer reliability. Clarke technique of ultrasound assessment has been claimed to reduce this issue and has lower false-positive rate [59].

Table 6.

Statistical comparison of different methods of neonatal hip screening for DDH

	Presence of risk factor for DDH	Clinical examination (combined Barlow and Ortolani test)	Ultrasound examination (combined static and dynamic)
Sensitivity (%)	10 [22]	62.59 [8], 100 [22], 18.5 [26], (mean 60.36)	75.78 [8], 100 [22], 47.6 [26], (mean 74.46)
Specificity (%)	98.1 [22]	99.76 [8], 88.9 [22], 99.6 [26], (mean 96.08)	99.85 [8], 98.5 [22], 99.6 [26], (mean 99.31)
Positive predictive value (%)	0.9 [22]	26.18 [8], 1.6 [22], 4.0 [26], (mean 10.59)	55.08 [8], 50 [22], 16.1 [26], (mean 40.39)
Negative predictive value (%)	99.8 [22]	99.9 [8], 100 [22] (mean 99.95)	99.9 [8], 100 [22] (mean 99.95)

Eight out of thirteen studies in current review favoured use of combined universal clinical neonatal and selective USG screening. It effectively reduced incidence of LDD when compared to universal clinical screening alone [16, 18, 23]. However, Rosendahl et al. [34] found no significant difference when clinical screening was performed by experienced and well-trained physicians. When compared to universal USG screening, rate of LDD in selective USG group was more but was not statistically significant [38]. This method therefore can be more suitable for communities where cost and resource availability is an issue. The Indian experience for selective USG screening in a retrospective cohort of 23,925 term newborns has been encouraging [60]. Selective USG screening gives a low primary treatment rate when compared to universal USG screening (Fig. 2) [16]. Simultaneously, a number of surgical interventions for late presentations are decreased leading to reduced overall cost of diagnosis and management [16, 41]. However, similar rate of late-presenting DDH was reported in England before and after introduction of selective screening programme [6, 15]. To improve the outcome and balance cost-effectiveness, repeat clinical examination beyond neonatal age was advised. American Academy of Pediatrics [61] further advised for a mandatory referral of children with positive clinical findings to an orthopaedic surgeon. Rebello et al. [45] suggested to review dislocatable hips clinically at 2 and 6 weeks and further clinical screening of all children brought to immunization centers. Talbot et al. [15] reported late presentations despite adding clinical evaluation at 6 week and suggested for more screenings in infancy. They associated the initial screening done by less experienced staff with increased incidence of late presentation and recommended that an experienced physician should do the screening. The ongoing selective screening programme in Scotland was upgraded with enhanced detection services [62]. A trained personnel (physiotherapist/physician) was appointed in postnatal ward with responsibility of training and supporting paediatric and maternity staff in doing hip examination, increasing awareness about DDH, and running weekly clinic for secondary screening. This strategy effectively reduced the risk of first surgery for DDH.

Favourable results of universal USG screening were reported by five out of six studies in current review [27–31]. It was more effective in decreasing incidence of LDD when compared to clinical screening alone [27, 29, 30]. Mark et al. [31] and Biedermann et al. [29] found no case of late diagnosis and number of operative procedures decreased by 75.9% in Austria. However, it has not completely eliminated late presentation in other studies [27, 30]. Possible reasons could be technical errors in performing scan or misinterpretation of USG finding. In spite of a right diagnosis, DDH may be inadequately treated or results of treatment may not be satisfactory leading to late presentation. Moreover, onset of DDH itself may be late [63]. Universal USG screening has also been reported to have significant false-positive rate leading to overtreatment [35, 40]. Unnecessary treatment in these otherwise normal children expose them to iatrogenic complications, increases parental anxiety, and cost of treatment. Rosendahl et al. [34] observed a higher negative predictive value for USG but suggested for refining the criteria for initiating therapy to avoid overtreatment. It remains a matter of debate due to economic concerns. It also needs equipment and expert radiologist which may not be readily available and is an important concern in many countries. Furthermore, spontaneous resolution of neonatal hip dysplasia detected on USG has been reported [64]. Olsen et al. [32] reported doubling of treatment rate with universal USG screening without any significant decrease in incidence of late presentation and recommended against it. In current review, 6 out of 11 comparative studies reported universal USG to be cost-effective with better detection rate and supported its use [33, 36, 40–43]. A delayed scan at 2–3 months was suggested to avoid overtreatment [40]. Three comparative studies advised against it [34, 35, 38]. Holen et al. [38] did not find any significant difference in incidence of late presentation when compared with selective USG screening and stressed for a high-quality clinical screening and selective use of USG. Geitung et al. [35] did not find the universal USG screening programme cost-effective when combined with universal clinical screening. However, they observed it to be cost-effective if clinical screening was omitted and ultimately recommended for selective USG screening. At an international interdisciplinary consensus meeting on evaluation of DDH organized in 2018, use of universal USG screening was firmly supported [65]. It was observed that USG screening reduces long-term consequences, is cost-effective, and does not cause overtreatment. Early screening within 6 weeks was recommended and Graf technique was suggested as technique of choice [65].

Cost of running a national screening programme can be a concern particularly in developing nations. It includes expense of training the physician, fee of physician, neonatologist working hour, and cost of ultrasound machine. Expenses involved in DDH care include cost for conservative and operative treatment including splints. Benefit is represented by expected treatment cost at a later date due to missed diagnosis at neonatal age [66]. Cost of surgical intervention is far more when compared to conservative management. An early diagnosis decreases number of late-presenting DDH and cuts down cost of surgical intervention. Economic justification of universal clinical neonatal screening against no screening is well established [33]. However, superiority and cost-effectiveness of universal clinical screening over selective or universal USG screening is debatable. Clegg et al. [36] reported comparable expense of overall management after comparing clinical screening alone with those with addition of selective and universal ultrasound. They recommended for universal USG screening after observing statistically significant reduction in number of surgical interventions per year and mean age at first operation. Holen et al. [38] compared selective and universal USG screening and observed same incidence of late-presenting DDH. They suggested for selective USG screening as it would be less expensive. Bralic et al. [66] did cost–benefit analysis of managing DDH at different ages of presentation, thereby evaluating economic justification for universal USG screening in Croatia. They observed cost difference of US$195,337 in treating DDH at 1 month and 4 months of age. This was 1.6 times the total cost of USG screening of all live birth that year. Furthermore, treatment cost using hip replacement in 165 missed cases alone would have covered total screening cost in entire country. They reported universal USG screening to be highly cost-effective and recommended for its nationwide implementation. A retrospective study from Austria compared the cost-effectiveness of clinical screening alone with universal ultrasound screening and reported a significantly high overall cost of treatment with clinical screening alone [42]. Introduction of ultrasound decreased the rate of splintage in infants and rate of surgical procedures in children over 18 months of age. They also observed that use of ultrasound reduces uncertainty and parental anxiety and may further reduce the cost to health services and family. It was suggested that a recommendation for selective USG screening by earlier studies was due to their failure to acknowledge intermediate findings like number of additional clinical follow-up and number of splints used and due to including patients in learning phase of ultrasound use. In a similar study, Gray et al. [41] reported lower cost of splinting and surgical intervention in universal ultrasound group and concluded against increased cost burden by the use of universal ultrasound method.

For a disorder to undergo nationwide screening, it should have a significant incidence rate and a known natural history, and should be best treated early before usual age of clinical presentation without any screening. Early treatment of DDH is easy to practice, and is known to change its natural history and indications for doing so are well defined. With an incidence up to 9.2 per 1000 live births, DDH in India needs a nationwide screening programme [4, 5, 60]. As discussed earlier, ideal screening method for DDH is still debatable. Literature supports universal USG screening and has proved its cost-effectiveness. However, considering the logistic and financial challenges in our country like shortage of well-trained radiologists and lack of resources, immediate implementation of universal ultrasound screening seems impractical. There are not any current guidelines for DDH screening in India. The time is ripe for professional organizations involved in the care of children and public health policy-makers to come together to develop national screening guidelines for DDH. The authors hope that the current systematic review of literature from countries which have well-established DDH screening programmes may assist them in formulating guidelines which are contextually relevant, practical, and cost-effective for the existing healthcare system in India.

Author Contributions

RAP: data acquisition and data analysis, drafting the manuscript, and proof reading. ANJ: conceptualized the study and made the study design, drafting the manuscript, and proof reading. Both authors read and approved the final draft of manuscript.

Declarations

Conflict of Interest

The authors declare that they have no conflict of interest.

Ethical Standard Statement

This article does not contain any studies with human or animal subjects performed by the any of the authors.

Informed Consent

For this type of study, informed consent is not required.