Determining clinical practice of expert physiotherapy for patients undergoing lumbar spinal fusion: a cross-sectional survey study

In this cross-sectional survey study, we asked expert physiotherapists who perform inpatient treatment before and after LSF to complete a survey on their practice routines. To select the expert physiotherapists, we contacted all heads of physiotherapy departments who were registered with the Dutch Association for Physiotherapy in Hospitals (NVZF) by e-mail (02/06/2014). The NVZF represents 67 general hospitals, academic hospitals and specialised care centres in the Netherlands [6]. Hospitals where LSF was not performed were excluded. Department heads were informed about the content of the study and were asked to forward the survey to their expert physiotherapist concerning LSF (i.e. the physiotherapist they would want to be treated by if they underwent LSF). Return of the questionnaire was considered as informed consent. A reminder was sent after 1 month. This manuscript is reported according to the STROBE guideline for cross-sectional studies [7] and the CHERRIES checklist for reporting the results of internet E-surveys [8]. Assessment by a medical ethics review board was not necessary.

The survey comprised 46 questions (nine open and 37 multiple-choice) on four domains: (1) demographic data (nine questions), (2) preoperative diagnostics and treatment (seven questions), (3) postoperative diagnostics and treatment (26 questions), and (4) information for discharge (four questions). The questions in the survey were based on a similar study in the UK by Rushton et al. [4]. However, we adapted the survey to the Dutch healthcare context. Moreover, we based the answer options for the questions on diagnostic procedures on the ICF core set for LBP [9]. Finally, we added 17 questions in order to obtain information regarding multidisciplinary cooperation, discharge criteria and referral information after discharge. The survey (translated into English; i.e., not an official cross-cultural adaptation) is available as an appendix to this manuscript.

To collect the data, we used Qualtrics (http://​www.​qualtrics.​com), a commonly used internet-based program for administering surveys [10]. To minimise the chance of incomplete responses due to skipped and/or forgotten questions, the function “Force Response” was used. “Skip Logic” was added to increase the efficiency of the questionnaire (completion time was approximately 15 min). Respondents were able to review and change their given answers using a back button. To prevent multiple answers from the same individual we checked from which hospital the questionnaire originated and their IP-address. In the case of duplicate entries, only the first entry was kept for analysis. The IP-addresses were deleted before the data was analysed. The questionnaire was pre-tested by five peers.

First of all, two researchers (ES and EJ) categorised the answers of the open questions and labelled them. Differences in categories between the two assessors were resolved by a third researcher (TH). In case of disagreement, the respondent was re-approached for further clarification. All data was analysed anonymously and presented as such. Completeness of the questionnaire was checked; forms were not included in the analysis if over 50 % of the data was missing.

Descriptive statistics were used to describe the study population and the survey answers [i.e. numbers and percentages, means and standard deviations (SD), and medians and interquartile ranges (IQR)]. To determine the level of uniformity between the physiotherapists on relevant goals and interventions used in expert standard practice, we used categories ranging from No uniformity to Very strong uniformity (see Table 1). Uniformity shows the percentage of participants choosing one answer option (per question). The more participants choosing one answer option, the higher the level of uniformity for a specific goal or intervention.

The majority of the respondents, representing 22 hospitals (79 %), did not provide preoperative physiotherapy care for patients undergoing LSF. In the six cases where preoperative care was provided, it was mainly group-based (five respondents) and aimed at informing the patient about the postoperative phase (six respondents). Regarding preoperative diagnostics and instructions there was no uniformity or low uniformity in goals and interventions such as performing a preoperative functional assessment (67 %), instructing patients on how to perform postoperative transfers (50 %) or taking a history (67 %) (see Table 3).

Inpatient physiotherapy for patients recovering from LSF was standard care in 22 hospitals (79 %). In four (14 %) cases, patient-specific needs were first assessed to establish the necessity of inpatient physiotherapy. In the remaining two cases it was unclear how physiotherapy was initiated and under which circumstances it was provided. Commonly, patients were treated once (50 %) or twice (34 %) a day by a physiotherapist for an average (SD) of 3.8 (1.3) days and a median (IQR) of 20 (18–25) min per session. In the majority of the hospitals (61 %) mono-disciplinary care was provided (i.e., by the physiotherapist). In the case where multidisciplinary treatment was reported in hospitals, the other professions most frequently involved were: family caregivers (11 %), nurses (11 %), and occupational therapists (7 %).

Respondents agreed to a great extent on the goals and interventions that are part of the inpatient rehabilitation process after LSF surgery (Table 4). Postoperative goals with strong to very strong uniformity were: getting the patient to function safely (93 %), getting the patient out of bed (93 %), informing the patient on the rehabilitation process (93 %), getting the patient to walk (96 %) with an optimal gait pattern (93 %), getting the patient to climb stairs (89 %) and getting the patient to carry out (bed) transfers (89 %). Postoperative interventions with very strong uniformity were: getting out of bed (93 %), walking (96 %) and climbing stairs (93 %). Interventions with strong uniformity were: taking patient history (86 %), giving advice on functional activities and restrictions (89 %), instructing and training patients about transfers (89 %), answering questions (89 %) and giving instructions for exercises at home (82 %).

No uniformity or low uniformity among respondents was seen for several goals and interventions, including: performing a physical examination (57 %), instructing how to lift and carry objects (32 %), and instructing how to use the restroom (32 %). Moreover, there was no uniformity regarding when to resume the activities of daily life after discharge (Table 5). Finally, a minority of respondents used questionnaires (4 %) or observational measurements (18 %) to guide or evaluate their therapy during the inpatient rehabilitation.

Of the responding physiotherapists 32 % always referred patients for outpatient physiotherapy and 50 % only if deemed necessary. Typically, according to the respondents, the decision to refer depends on the patient’s physical capacity, coping ability or on the physician’s advice. Our respondents stated that the majority of patients are discharged to their home with a referral to primary care physiotherapy (78 %). At which practice the rehabilitation process after discharge is continued is mainly decided by the patient (67 %) and primarily based on the distance from their house to the practice (67 %).

A number of strengths and weaknesses are apparent in this study. The strengths include the validity of our findings regarding the Dutch inpatient physiotherapy practice, as 74 % (67 out of 91) of the Dutch hospitals [6] were approached with a response rate of 93 %. Furthermore, we focused only on expert physiotherapists, allowing an overview of expert based care as reported both before and after LSF. We believe we were successful in doing so, as the majority of physiotherapists had 8 or more years of experience treating patients undergoing LSF.

Some limitations include the external validity of our findings. After all, the data might not be generalisable to some other countries due to differences in cross-cultural health care, educational systems and curricula, although the findings might be relevant to other European countries due to their similarities in culture and healthcare systems. Furthermore, questionnaires rely on self-reported data and therefore do not guarantee an accurate reflection of daily clinical practice. Ideally, observations of clinical practice would be performed; however, due to time and budgetary constraints, this method was not feasible. Finally, we aimed to include expert physiotherapists through asking the department heads to select the therapist who they would want to be treated by after an LSF procedure. A better method for selecting experts would be based on their clinical outcomes; unfortunately this data is not available [11].

There is just one other study that describes the current practice of physiotherapy for patients undergoing LSF [4]. This study investigated physiotherapy practice for patients undergoing LSF in the UK [4]. The authors administered a nationwide survey, targeting all physiotherapists that were involved in the management of patients before or after LSF within the UK National Health Service trusts. Our findings overlap considerably with theirs. For instance, in both the UK and the Netherlands: (a) physiotherapy care is provided structurally after surgery (70 vs. 79 %, respectively); (b) few centres used questionnaires and performance measures to evaluate or monitor the treatment (6 and 19 vs. 4 and 18 %, respectively); and (c) interventions such as providing information (98 vs. 89 %, respectively), answering questions (95 vs. 89 %, respectively), and instructing and supervising walking (98 vs. 96 %, respectively) were most common.

A notable difference between practice in the Netherlands and the UK is the use of therapeutic protocols; 49 % in the UK and 82 % in the Netherlands. In the Netherlands it is common for hospital physiotherapists to protocolise their postoperative care [12]. Through the use of protocols, physiotherapists who are unfamiliar with working practices in other departments, can still deliver care as best as possible. Unfortunately, these protocols carry the risk that all therapists will deliver protocolised and therapist centred (one-size-fits-all) care instead of the currently favoured patient centred and personalised care [13, 14], as demonstrated by the high number of expert therapists using protocols to guide their day-to-day practice therapy in our study. We specifically included expert physiotherapists in our survey population to distil best (physiotherapy) practice in LSF [13, 14]. Interestingly, we found that factors essential for clinical reasoning (such as functional diagnosis) were often not evaluated, therapy was either never or always provided (regardless of the patient’s need), and therapy was typically delivered on time-based principles (not goal-based). It seems that now is the time for best practice guidelines to be established.

In the Netherlands it may be necessary to reconsider the approach of the (expert) hospital physiotherapist in the management of individuals undergoing LSF. Considering that: (1) there is a (small) number of hospitals where LSF surgery is routinely performed without involving physiotherapists in the clinical care pathway, possibly due to the lack of evidence on benefits of clinical physiotherapy after lumbar surgery [14]; (2) physiotherapy in the management of LSF is mainly characterised by one-size-fits-all care, rather than care based on evaluating the specific functional needs of the patients [15, 16]; and (3) there is no uniformity or low uniformity in the different aspects of the content of the physiotherapy management (e.g. the necessity of preoperative care, multidisciplinary treatment and the contents of advice); the current physiotherapy practice needs to be reconsidered.

A shift from postoperative care to preoperative care in patients undergoing major surgery and at-risk for poor outcomes could decrease costs, improve functional outcomes, and in some cases, prevent complications and death [17]. This may hold true for individuals undergoing LSF surgery as well [18, 19]. Preoperatively predicting which patients will not benefit from LSF has proven to be quite challenging, as most medical and surgical factors have very little predictive value [20, 21]. Our data demonstrates that pre- or postoperative risk-stratification and/or optimisation are not utilised in daily clinical practice. Nonetheless, evidence tells us that functional measures are vital for risk assessment and provision of optimal care before and after major surgery [22‐24].

An interesting finding was that there is little agreement on the “dos and don’ts” after LSF surgery. Even though nearly all physiotherapists report that they provide information and recommendations, we found there is not only little uniformity in the content of recommendations but also in when to resume the activities of daily life. Topics that are almost never discussed are: (when to) return to sports and (when to) return to work, despite these being absolutely crucial for reintegration and participation in society. This apparent dissensus among health professionals on the timing of postoperative activities might be caused or at least maintained by the scarce, and somewhat counter-intuitive, literature on this topic [25, 26].